8fer- ~ (4)3 ~S)J Tte WI6+ MAampr-ialis J 2(110 bull

1 Against Materialism

Laurence Bonjour

Recent philosophy of mind has been dominated by materialist (or physicalist) views views that hold that mental states are entirely material or physical in nature and correlatively that a complete account of the world one that leaves nothing out can be given in entirely materialist terms Though (as the tirle of this volume suggests) this may be changing to some extent philosophers ofmind who are willing to take seriously the possibiliry that materialism might be false are still quite rare

I have always found this situation extremely puzzling As far as I can see materialism is a view that has no very compelling argument in its favor and that is confronted with very powerful objections to which nothing even approaching an adequate response has been offered The central objection elaborated in various ways below is that the main materialist view quite possibly the only serious materialist view offers no account at all of consciousness and seems incapable in principle of doing so But consciousness as Nagel pointed out long ago2 is the central feature of mental states-or at the very least a feature central enough to

make a view that cannot account for it plainly inadequate Supposing as I will try to show below that this assessment is correct

why have materialist views been so dominant Part of the answer is that it is far from clear that dualist views at least those that go much beyond the bare denial of materialism are in any better shape (see the last section of this chapter for some elaboration of this) But it must be insisted that the inadequacies of dualism do not in themselves constitute a strong case for

Admittedly the rather stark simplicity of this formulation is not to evetyones taste There are those who prefer a formulation in terms of the superomimce of the mental on the physical and some (one variant of the recendy popular view known as non-reductive materialism) who want to interpret such a formulation in a way that gives the mental some sort of ontological independence (For a recent discussion see Antony (2007) I have no space here to sort out the twistings and turnings of that discussion and will simply assume that any genuine ontological independence of the mental would amount to a kind of epiphenomenalism (see further below) rather than genuine materialism and thus that a genuine materialism must be committed to the formulation in the text

2 In his classic 1974 paper What Is It Like to Be a Bat Philosophical Review 83435-50

4 5 L Bonjour

materialism arguments by elimination are always dubious in philosophy and never more so than here where the central phenomenon in question (that is consciousness) is arguably something of which we still have little if any real understanding Instead materialism seems to be one of those unfortunate intellectual bandwagons to which philosophy along with many other discipshylines is so susceptible-on a par with logical behaviorism phenomenalism the insistence that all philosophical issues penain [0 language and so many other views that were once widely held and now seem merely foolish Such a comparison is misleading in one important respect however it understates the fervency with which materialist views are often held In this respect matershyialism often more closely resembles a religious conviction-and indeed as I will suggest further in a couple of places below defenses of materialism and especially replies to objections often have a distinctively scholastic or theological flavor

In what follows I will try to substantiate this indictment of materialism by doing the following things First I will look at some of the main considerations that are advanced in favor of materialism in general as opposed to particular materialist views attempting ro show that these are surprisingly insubstanrial and rest mainly on assumptions for which no real defense is offered Second I will look at the overwhelmingly dominant materialist view namely functionalism arguing that it is deeply inadequate in relation ro the problem of consciousness Third I will look at what is widely regarded as the most serious specific problem for materialism in general and functionalism in particular namely the problem of qualitative content or qualia focusing here on a somewhat modified version of Frank Jacksons well-known knowledge argument and trying to show that the objection ro materialism that results is still extremely compelling Fourth I will look at a problem that functionalism is claimed ro handle more successfully the problem of intentional states arguing that there are dear cases of conscious intentional states which materialism in general and functionalism in particular can handle no better than qualia-and for essentially the same reasons Fifth and last I will ask what lessons if any for a more adequate account of conscious mental states can be derived from all of this

1 THE CASE FOR MATERIALISM

One of the oddest things about discussions of materialism is the way in which the conviction that some materialist view must be correct seems to

float free of the defense of any particular materialist view It is very easy to

find people who seem to be saying that while there are admittedly serious problems with all of the specific materialist views it is still reasonable to preshysume that some materialist view must be correct even if we dont know yet

igainst Materialism

which one or that the seeming force of the objections to particular matershyialist views must be balanced against the strength of the underlying case for materialism But why is this supposed to be a reasonable stance to take What arguments or reasons or otherwise compelling intellectual considerations are there that could yield a strong background presumption of this sort in favor of materialism (or create a substantial burden of proof for opponents of materialism)

There are of course arguments against particular versions of dualism mainly against the imeractionist version of Cartesian substance dualism For reasons already mentioned I will set these aside as not constituting in themselves an argument for materialism There is also the inductive generalization from the conspicuous success of materialist science in a wide variety of other areas This undeniably has some modest weight but seems obviously very far from being enough to justifY the strong presumption in question Inductions are always questionable when the conclusion extends to cases that are significantly different from the ones to which the evidence pertains and even most materialists will concede that conscious phenomena are among the most difficult-indeed seemingly the most difficult ofall-for materialist views to handle Thus the fact that materialism has been successful in many other areas does not yield a very strong case that it will succeed in the specific area that we are concerned with

Beyond this there seem to be only rwo related sorts ofgrounds that are offered for a Strongly pro-materialist presumption both of which are quite flimsy when subjected to any real scrutiny

1 L The Principle of Causal Closure

The first and clearer of these rwo grounds appeals to the thesis that the material universe is cl1usI1L~l claud that material things are never causally affected by anything non-material (so that as it is often put physical science can in principle give a completely adequate explanation of any physical occurrence without needing to mention anything non-physical) This thesis is commonly referred [0 as a principle a characterization that leaves its status rather obscure (Philosophers often seem to describe something as a principle when they are inviting their readers to accept it as a basis for further argument even though no clear defense of it has been offered)

I The closure principle does not by itself entail that materialism is true It leaves open both the possibility of non-material realms that are causally isolated from

I the material world and also the possibiliry that epiphenomenalism is true that conscious phenomena are side-effects of material processes that are incapable of having any reciprocal influence on the material world But assuming that the non-material realm in the first possibiliry is supposed to be the locus of conscious phenomena both of these possibilities are extremely unpalatable even paradoxical in essentially the same way The main problem is not as

i

7 6 L Bonjour

is often suggested that they are incompatible with the general common sense intuition that conscious states causally affect bodily behavior A more specific and serious problem is that if either of these possibilities holds then it becomes difficult or seemingly impossible to see how verbal discussions of conscious phenomena-such as this chapter and many others-can be genuinely about them in the way that they seem obviously to be How can people be talking about conscious states or saying anyrhing significant about them if completely adequate causal explanations of their verbal behavior can be given that make no reference to such states Even without invoking any specific version of the causal theory of reference it is hard to see why verbal discussions that are entirely unaffected by what they purport to be about should be taken seriously Thus while a number of philosophers have in recent times been seemingly tempted by epiphenomenalism it appears that they can have been genuinely advocating such a view about conscious states only if the view itself is false 3

For these reasons the argument from the principle of causal closure to the truth of materialism is quite strong even if not fully conclusive But why is the principle of causal closure itself supposed to be so obviously correct Clearly this principle is not and could not be an empirical result no empirical investigation that is at all feasible (practically or morally) could ever establish that human bodies the most likely locus of such external influence are in fact never affected even in small and subtle ways by non-material causes Weare told that scientists accept this principle and often that most philosophers accept it as well But do they have any compelling reasons for such acceptance Or is this vaunted principle nothing more than an unargued and undefended assumption-a kind of intellectual prejudice in the literal meaning of the word

Taken in the abstract apart from any appeal to a specific account of conscious mental phenomena I have no idea whether the principle of causal closure is true or not More importantly I cannot imagine how to rationally decide whether it is true without first arriving at a defensible account of conscious mental states It seems utterly obvious that mental states do causally affect the material realm probably by causally affecting the actions of human bodies in general but (as just argued) at least more narrowly by causally affecting verbal discussions of these matters If a materialist account of conscious states is correct then the principle of causal closure seems likely to be true But if no such account is correct then the principle is almost certainly false Thus to argue for the truth of materialism or for a strong presumption in favor of materialism by appeal to the principle of causal closure is putting the cart in quite a flagrant way before the horse

3 This problem seems ro be the main reason for Jacksons abandonment of his previous antishymaterialist stance (Jackson never rook seriously the possibility that the non-material qualia for which he was arguing might causally affect the material world)

Against Materialism

12 The Appeal to Naturalism

A second sort of defense of a general presumption in favor of materialism appeals to the general idea of naturalism Here again we have a view like materialism itself to which many many philosophers pay allegiance while offering little by way of clear argument or defense but here the view itself is much harder to

pin down in a precise way Indeed even more striking than the absence of any very clear arguments is the fact that many recent philosophers seem so eager to commit themselves to naturalism-to fly the naturalist flag as it were-while showing little agreement as to what exactly such a commitment involves Thus naturalism seems to be even more obviously an intellectual bandwagon than materialism (In addition naturalism for some of those who use the term seems to just amount to materialism which would make an argument from naturalism to materialism entirely question-begging)

Is there any genuine support for a materialist presumption to be found in the viciniry of naturalism One version of naturalism is the idea that metaphysical issues-or philosophical issues generally-should be dealt with through the use of the methods ofnatural science If this is accepted and ifit is true that following the methods of natural science leads plausibly to an endorsement of materialism then at least some presumption in favor of materialism might follow But both of the needed suppositions are in fact extremely dubious to say the least There is simply no good reason to think that the methods of natural science exhaust the methods of reasonable inquiry-indeed as has often been pointed out there is no plausible way in which that claim itself can be arrived at using those methods Nor is there any very clear reason to think that applying the methods of natural science to the question of whether materialism is true assuming that one could figure out some reasonably clear way to do that would lead to the conclusion that materialism is correct Such a conclusion is obviously not within the purview of physics but it is also not within the purview of psychology especially as currently practiced As was true with closure there is no doubt that many (but not all) natural scientists assume the truth of materialism but the question is whether they have any good reason for such an assumption-a reason that would itself have to transcend their strictly scientific claims and competence 4

Thus while the murkiness of the discussions of naturalism makes it harder to be sure naturalism like closure does not seem to yield an independently defensible presumption in favor of the truth of materialism If there is any better

bull Lurking here is the difficult issue of what sons of entities or properties count as material or physical Is there any good war ro delimit the realm of the material that does not preclude further discoveries in physics but also does not trivialize the categoty by allowing it to include anything that people in departments labeled Physics might eventually come to study This IS anything but a trivial problem but I have no space here to pursue it further

8 9 L Bonjour

rea~n or basis for such a presumption that is prior to and independent of the defense of some particular materialist view I have no idea what it might be

2 FUNCTIONALISM AND CONSCIOUSNESS

The upshot of the previous section is that the case for materialism must rest almost entirely on the defense of particular materialist views and not to any substantial extent on any background presumption So what materialist views are there The answer I think is that once both logical behaviorism and various versions of e1iminativism are set aside as toO implausible to be taken seriously-something that I will assume here without any further discussion-there is only one main materialist view namely functionalism with no very serious prospect that any others will emerge And the fundamental problem for materialism I will suggest is that functionalism offers no account at all of consciousness and seems in principle unable to do so

What gives rise to the mind-body problem in the first place and poses the essential problem that any adequate version of materialism mUSt solve is the fact that conscious mental states as we are aware of them do not present a material appearance-do not seem as we experience them to be material in their makeup in any apparent way Thus a view which holds that everything that exists is material must either (a) deny the very existence of such states as e1iminativism does or else (b) explain how states and correlative properties that do not initially seem to be material in nature can nevertheless turn out to be so A view that takes the latter alternative must give an account of the nature of such states and properties that both accurately reflects their character as experienced and explains how they can nonetheless be entirely material in their makeup And this I suggest is something that has never been successfully done

The starting point for modern versions of materialism was the central-state identity theory particularly the version advocated in a famous paper by J J c Smart (1959) Smart recognized that the truth of materialism can only be an empirical discovery not something knowable a priori For this to be so he argued the various mental states in question must be conceived in a topicshyneutral way a way that makes it possible for them to be merely material in character without implausibly requiring that this be so Only in relation to such a conception would it be possible to discover empirically that such a state is in fact a neurophysiological state of some kind

But for this to work it is crucially important that the topic-neutral conshyception in question be adequate to capture the essential features of mental

Against Materialism

states-something about which Smart was less clear than he might have been For only if this is so will it be the case that showing that the conception offered can be realized by a material state can establish that mental states might in fact be merely material states thereby allowing the rest of the argument to proceed on grounds of simplicity as Smart suggests Whereas if the proposed topic-neutral conception leaves out essential features of mental states-such as consciousness-then the mct that material states can satisfy that conception will be insufficient to explain how mental states might just be material states (Smarts own attempt at a topic-neutral characterization fails to distinguish conscious mental states from whatever else might be going on in the person under a particular set of circumstances)

As in Smarts view functionalism in effect attempts to offer a topic-neutral characterization ofmental states one which will allow but not require that they be essentially material in characters The more general functionalist characterization is in terms of causal role a mental state is characterized by its causal relations to sensory inputs behavioral outputs and other mental states of the same sort The functionalist then proceeds to argue that the states thus characterized could perfectly well be material srates even though the functional characterization does not require rhis A further widely discussed aspect of the view is that different material states could satisfy the functionalist characterization of a particular mental state in different SOrtS ofcreatures or even in the same creature at different times so that (on the most standard version) a material state realizes a functionally characterized state but is not strictly identical with it

But the deepest problem for the functionalist is that the characterization of mental states in terms of causal role say~ nothing at all about consciousness or conscious character There is no apparent reason why a stare thar realizes a particular causal role would thereby need to have any specific sort of conscious character (the point made by the familiar reversed spectrum cases)-or indeed any conscious character at all Thus to point out that a physical state could realize such a causal role really does nothing at all to explain how a conscious state could be (or be realized by) a merely physical state In this way functionalism fails utterly to offer any explanation or account of the most important and conspicuous feature of mental states-or at the very least of a very important and conspicuous feature

It may seem hard to believe that a view that has been held by so many people for so long can be so easily shown to be inadequate in a fundamental way but I think that this is nonetheless so The only solution would be

5 This way of looking at functionalism is explicit in David Lewiss discussion of one of the earliest versions of the view in his 1966 paper An Argument for the Identity Theory Journal of Philosophy 63 17-25 see p 20

10 L Bonjour Against Materialism 11

to offer some supplementary account of what material features give rise to

conscious experience But I know of no such account at least none with any real plausibiliry6

This difficulty with materialism in general and functionalism in particular has of course occasionally been recognized But it still seems to have had remarkably little impact on the widespread acceptance of materialist views I have no very good explanation to offer of this though part of the reason is perhaps the prevailing tendency to approach the philosophy of mind from a third-person neo-behaviorist perspective in which consciousness is largely or entirely ignored (But on this issue it is hard to distinguish the chickens from the eggs)

My basic case against materialism is complete at this point there is no good reason for any strong presumption in favor of materialism and the main materialist view fails to offer any real explanation of a central aspect of mental states namely their conscious character meaning that there is no good reason to think that it is correct as an account of such states But though this very simple argument seems to me entirely compelling I will elaborate it further in the next two sections by focusing on the two main specific kinds of mental states The version of the argument that applies to states with qualitative content is very familiar even though I think that its full force has still not been generally appreciated In contrast the application of essentially the same basic argument to conscious states with intentional content has received far less attention

3 THE PROBLEM OF QUALITATIVE CONSCIOUSNESS

MARY REDUX

Though functionalism fails to adequately account for consciousness of any SOrt

perhaps the most conspicuous aspect ofthis failure penains to qualitative content the sort of content involved in experiences of color and sound and of like pains and itches This point has been made in many ways but the most straightforward and compelling in my view is still the so-called argument initially suggested by Thomas Nagel in relation to the experiences of bats and later developed by Frank Jackson using his famous example of black-and-white Mary on which] will mainly focus here (As most will know

6 One possibiliry is the so-allIed higher order thought theory which holds that consciousne arises when one mental state is the object of a second higher order mental state I have no room here to consider this view in detaiL But the basic-and obvious-problem with such a view is that there is no reason why there could not be a hierarchy of sort even one with many more levels in which there was no consciousness involved at alL (For some elaboration see my contribution to

Bonjour and Sos (2003) pp 65-8) 7 See for example Colin McGinn (1989) David Chalmers (1995) and David Chalmers (1996)

Jackson has changed his mind about this argument and now rejects it though his reasons seem to me unpersuasive)

I will assume here that Jacksons original version of the saga of Mary is familiar enough to require only a brief summation Mary is a brilliant neurophysiologist who lives her entire life acquires her education and does all of her scientific work in a black-and-white environment using black-and-white books and blackshyand-white television for all of her learning and research In this way we may suppose she comes to have a complete knowledge of all the physical facts in neurophysiology and related fields together with their deductive consequences insofar as these are relevanr-thus arriving at as complete an understanding of human functioning as those sciences can provide In particular Mary knows the functional roles of all of the various neurophysiological states including those pertaining to visual perception by knowing their causal relations to sensory inputs behavioral outputs and other such states But despite all of this knowledge Mary apparently does nor know all that there is to know about human mental states for when she is released from her black-and-white environmenr and allowed to view the world normally she will by viewing objects like ripe tomatoes learn what it is like to see something red and analogous things about other qualitative experiences But then comments Jackson it is inescapable that her previous knowledge was incomplete But she had all the physical information Ergo there is more to have than that and Physicalism is falses

Despite the initial force of this rather simple argumenr materialists have nOt been persuaded and the literature comprising materialist responses to the Mary example is very large 9 One thing to say about these responses is that few if any of them are even claimed ro have any substanrial independent plausibility instead they are PUt forward in a way that takes for granted the sort of general presumption in favor of materialism and correlative burden of proof for antishymaterialist views that J have argued does not genuinely exist A full discussion of these responses is impossible here but there are some main themes that can be

dealt with in a general way 10 One of these is the suggestion that although something new when she leaves the black-and-white

room what she acquires is not a knowledge of a new foct (or facts) bur rather something else A second is the suggestion that what she does something like a new ability perhaps more specificall representational ability And if these rwo themes are

bull Frank Jackson (1982) p 130 see also Frank Jackson (1986) bull Many of these discussions are collected in Ludlow Nagasawa and

)0 For a useful taxonomy of the various possible materialist responses see Robert Van Gulick So Many Ways of Saying No to Mary in Ludlow Nagasawa and Stoljar (2004) pp 365-405 (This is of the places where the materialist discussion bears a striking similariry to scholastic theology one can easily imagine a complacent theist writing an artide entitled So Many Ways to Answer the Problem of Evil)

12 L Bonjour i Against Materialism 13

fresult is that there is nothing about the Mary example that is incompatible with Though we have made things vastly easier for Mary by focusing on twO cases the truth of materialism involving colors with which she is now familiar rather than asking her to figure

Particularly in light of rhe general materialist failure to provide an account out on the basis ofher overall physicalneurophysiologicalfunctional knowledge I of conscious experience I doubt very much whether any response of this son would seem even mildly convincing to anyone who was not already determined to adhere to materialism come what may But the first of these fWO themes does at least point to a kind of lacuna in Jacksons original account of the case if Mary learns new facts what exactly are they Indeed in his response to an early version of this suggestion Jackson is reduced to invoking the problem of other minds as a (very) indirect basis for thinking that factual knowledge of some son is involved II

It is however surprisingly easy to modifY the original case in a way that makes it utterly dear that there are facts that Mary does not know while she is in the black-and-white room and will learn when she emerges Suppose that while she is still in the otherwise black-and-white environment fWO color samples are brought in one a sample of a fairly bright green approximately the color of newly mown grass and the other a sample of a fairly bright red approximately the color of a fire engine Mary is allowed to view these samples and even to know that they are fWO of the colors that she has learned about in her black-and-white education She is not however told the standard names of these colors nor is she allowed to monitor her own neurophysiology as she views them

We now remind Mary of two specific ca~es that she has studied morougIllV and about which she knows all the physic~l One of these is a case where a person was second is a case where a person was lookmg at a tell Mary that one of these people had an experience predominar one of the colors with which she is now familiar and that the other person had an experience predominantly involving the other color but of course not which was which If we call the colors presented by the samples color A and color B Mary now knows that one of the two following pairs of claiflls is true

(l) The experience of freshly mown grass predominantly involves color A and the experience of a newly painted fire engine predominantly involve~ color B

(2) The experience of freshly mown grass predominantly involves color B and the experience ofa newly painted fire engine predominantly involves color A

But can she tell on the basis of her black-and-white knowledge together with her new familiarity with the two colors whether it is (1) or (2) which is true (Notice carefully that there is no apparent problem with her understanding of these claims)

l See Jackson (1986 294)

what color experiences in general are like it still seems quite dear for essentially the same reasons that were operative in the original case that she will have no more success with her much more limired task Just as there was nothing in the physical account that could tell her what an experience of red was like so there is still nothing in the physical account of the fire engine case that could definitively pick out the color of one of the samples as opposed to the other 12

And yet whichever of (1) and (2) is true states a foct (or facts) in as robust a sense as one could want-a fact that Mary will learn when she emerges from the black-and-white room and is allowed to view ordinary objects of various kinds

Moreover if there are abilities that result from experiencing the two colors in question Mary presumably can acquire them on the basis of the samples Consider for example Harmans suggestion l3 that what Mary acquires in the original case when she leaves the black-and-white room and sees red for the first time is a perceptual concept of red one that essentially involves being disposed to form perceptual representations involving it in the presence ofcausal stimulation of the right SOrt-sO that she cannot acquire that concept in the original version of the black-and-white room There is much that is questionable about the idea ofsuch a concept but if there is indeed such a thing then Mary in the new version of the case presumably can acquire it by viewing the red sample

more than one sample is for some reason required but it would be easy the new version of the case to allow for that) So we may

suppose Mary has the perceptual concept of red and the perceptual concept of green but she still cannot figure our from her physical knowledge which of these concepts is being employed by the people in the cases she has studied Yet this too is a fact and ifmaterialism isrrue an entirely physical fact So why cant she know it

Here as far as I can see there are only fWO possible moves for the materialist which are even marginally worth considering One is the suggestion that Mary already knows the facts in question as a part of her overall physical knowledge but that she knows them under a different guise or mode of presentation than that under which she will come to know them when she leaves the blackshyand-white room This idea can be developed in different ways and with enormous technical ingenuity But does it really have any serious plausibility Imagine that

J2 As Jackson emphasizes in Jackson (1986 295) it is nO enough for Mary (0 be able to conjecture or guess at the answerO this question For physicalism to be true the fact in question must actually be contained in her physical knowledge

U Gilbert Harman (1990) The Intrinsic Quality of Experience Philosophical Pmpectives 4 3J~52 at pp 44~5 Harman does not actually melllion the Maty case as sucn focusing instead on a person who is blind from birth but still learns all the physical and functional facts of color perception But he does cite Jackson (along with Nagel) as the source of tne objection he is discussing

14 15 L Bonjour

Mary in our modified version of the case having finally experienced real colors is eager to find out more about these intriguing features of the world about which she has been kept in ignorance She wants very much for example to know whether it is (I) or (2) that captures the relevant facts about cases of that sort-and is seriously frustrated about being kept in ignorance any longer Suppose that we respond to her frustration by informing her that she already knows the very facts that she is so eager to learn Surely she would not be satisfied How might she respond

I think we can imagine Mary saying something like this

You philosophers are really amazing The idea that I ar(atiy know the facts I am interested in-indeed all facts of that general kind-is simply preposterous I know all of the physical details but none of them tells me which of the properties I have just experienced on the basis of the samples is realized in each of the two cases If you suggest to me that there arent really novel properties but rather novel concepts or ways of representing or whatever then (while finding that suggestion itself pretry hard to swallow) I would still insist that which concept or way of representing is involved in each case is still something that my physical knowledge doesnt give me any clue about Perhaps as you say there is some clever or complicated way in which the things I want to know are related to the physical things I do know-maybe there is even some metaphysically necessary connection between them (assuming that it is kosher for materialists to believe in such things) Anything like that however just adds to the list of facts that my physical knowledge doesnt reveal to me I am a scientist and not a philosopher so Im not really sure which conception of a fact is the right one (All of the ones you suggest seem pretty weird) But there is undeniably something that I want to know-something that is true about the world-that cant be learned on the basis of all my physical knowledge And that means that the physical story isnt in fact the whole story

Not surprisingly I think that the response I have imagined for Mary is exactly right-that any way of understanding or individuating facts according to which some piece of Marys physical knowledge and either (1) or (2) above turn out to be formulations of the same fact is a conception of fact that is simply too intuitively implausible to be taken seriously

The other possible materialist response is to gram that Mary will learn new facts when she emerges from the black-and-white room in the modified version of the case but to insist that these are nonetheless still physical facts On this view what the case shows is that it is impossible for Mary to acquire complete physical knowledge in the black-and-white room One way to put it is to say that while she can learn all the objective physical facts there are still certain subjective physical facts l4 that she cant learn One can learn what it feels like subjectively to be an organism ofsuch-and-such a general physical description in such-and-such a specific physical state only by actually realizing that condition But that it feels

14 See Van Gulick (2004) for one version of this suggestion

Agaimt Materialism

a certain way or involves a certain sort of conscious experience is still on this view an entirely physical fact

I have to admit that I find it nearly impossible to take this response seriously The only argument for it seems to be an appeal to a background presumption in favor of materialism that is so strong as to make it allegedly reasonable to claim that any fact there is must be a material fact even ifwe cant see in any clear way how it could be a material fact Materialism as we have already seen offers no real account or explanation of consciousness and so also no reason for thinking that there is any subjective experience at all involved in being in a certain material state Thus to advance a view of this sort is in effect just to insist that no fact of any sort can be allowed to refute materialism and thus that any possibiliry of this sort must simply be absorbed into the materialist view however inexplicable in materialist terms it may be (It is not much of a stretch to imagine the materialist saying that we must first believe in order than we may understand)

Thus the modified version of the Mary case seems to present an objection to materialism in general (and functionalism in particular) that is about as conclusive as philosophical arguments ever get However exactly they should be characterized there are facts that Mary cannot know on the basis ofher complete physicalneurophysiologicalfunctional knowledge even when she is given the sort of limited experience needed to understand the claims in question and to acquire any abilities that might be relevant These facts do not seem to be material facts and there is no basis that is not utterly arbitrary and question-begging for supposing that they are Thus we have the strongest of reasons for holding that the materialist account of realiry is incomplete and hence that materialism is

false

4 THE PROBLEM OF CONSCIOUS INTENTIONAL

CONTENT

Qualia of the son involved in the Mary case are widely recognized to pose a serious problem for materialist views and it is not too hard to discern occasional misgivings in this respect under the faltade of materialist confidence But as already mentioned above it is widely assumed that materialism is in much better shape with regard to intentional mental states propositional attitudes and other states that involve aboutness I believe however that this is almost entirely an illusion-that the problems for materialism are just as serious in this area with consciousness being once again the central focus

Materialist accounts of intentional states tend to focus mainly on dispositional states such as beliefs and desires Given the central role of such states in explanations of behavior this is in some ways reasonable enough But such a focus tends to neglect or even ignore the existence of comcious intentional states-even though having conscious thoughts that P is surely one of the

16 L Bonjour

central things that having a dispositional belief that P disposes one to do A focus on belief in particular also has the unfortunate effect of making externalist accounts of intentional content seem more plausible than they possibly could if the emphasis were on conscious intentional states

For these reasons I will focus here explicitly on conscious thoughts As I sit writing this chapter a variety of conscious thoughts pass through my mind Many of these involve the assertion or endorsement of various propositions that materialism cannot account for consciousness that the trees outside my window are very bare that the weather looks cold and dank that the situation in the Middle East looks grim and so on Other thoughts are also propositional but in a way that does not involve assertion my conscious desire to get several pages written before lunch the hope that the stock market will continue to rise and so on It is doubtful that conscious thought must always be propositional in character but it will in any case simplify the issues to be discussed if we largely ignore the propositional aspect of these various thoughts and focus simply on their being conscious thoughts of or about various things or kinds of things materialism the trees the Middle East the stock market and so on

One crucial feature of such conscious thoughts is that when I have them I am in general consciously aware of or consciously understand or grasp what it is that I am thinking about (and also what I am thinking about it) When I think that the trees outside my window are bare I consciously understand that it is certain trees that I am thinking about (and along with this what sort of thing a tree is and which trees I have in mind) What exactly this conscious grasp of the object of thought involves varies from case to case and is sometimes not easy to precisely specify Moreover as will emerge it is something of which I think we presently have no real explanatory account of any substance But its existence is I submit completely undeniable Indeed being able in this way to consciously think about things to have them in mind is in many ways the most central and obvious feature of our mental lives

It is obvious that a persons conscious grasp of the object of their thought of what they are thinking about can vary on a number of dimensions it may be more or less precise more or less detailed more or less clear more or less complete But contrary to what is sometimes suggested it is rarely if ever merely disquotational in character Perhaps (though I doubt it) there are cases where a scientifically untutored person is thinking about eg electrons and where their sole grasp of what they are thinking is that it is what is referred to in their society or community by the word electron -so that what they are thinking about is in effect electrons (whatever they are) But this is surely not the ordinary situation when we think about various things 15

15 Notice that even a thought about what the relevant Ioci~tal ocp~rts mean by electrons would have to involve a non-disquotational element in the reference to those experts and also in the reference to the word to think about whatever it is that the societal experts mean by electrons

Against Materialism 17

Moreover the existence of conscious intentional content is perfectly compatshyible with the existence of an externalist dimension of thought content-though not with the view that all content is external If as in Putnams famous example a person is thinking about earthly water at a time prior to the discovery of its chemical composition there is no reason to deny that they are in a sense thinking about H 20 But in such a situation the aspect of being about H 20 will obviously not be part of their conscious internal grasp of what they are thinking about in the way that the more superficial aspects of water will 16 And in a somewhat parallel way the person in Burges famous example who thinks he has arthritis in his thigh and to whom our standard belief-ascription practices ascribe a belief about arthritis (where this is among other things a disease that only occurs in joints) obviously does not consciously grasp the disease that he is thinking about in a way that involves this specific feature of itY But it is nonetheless impossible to describe either example in a convincing way without presupposing that the people in question do have something consciously in mind a substance having the superficial properties of water in Putnams example and a disease having certain fairly specific features in Burges Thus while it is possible to dispute the relative importance ofconscious internal thought content and external thought content of which the subject is not conscious examples of this sort provide no basis at all for denying that conscious internal content exists

The issue I want to raise here is whether a materialist view can account for the sort of conscious intentional content just characterized Can it account for conscious thoughts being about various things in a way that can be grasped or understood by the person in question In a way the answer has already been given Since materialist views really take no account at all of consciousness they obviously offer no account of this particular aspect of it But investigating this narrower aspect of the issue can still help to deepen the basic objection to

materialism Here it will be useful to bring the brilliant neurophysiologist Mary briefly

back onto the scene even though the black-and-white aspect of her situation is no longer relevant Suppose that Mary studies me as a subject and comes to

have a complete knowledge of my physical and neurophysiological makeup as I am thinking these various thoughts Can she determine on that basis what I am consciously thinking about at a particular moment

One thing that seems utterly clear is that she could not do this merely on the basis of knowing my internal physical characteristics-as it is sometimes put

I knowing everything physical that happens inside my skin There is no reason t f

is not the same thing as thinking about whatever is meant by electrons by whatever is meant by the societal experts

I 16 See Hilary Putnam (1975a) (Cambridge Cambridge Universiry Press) pp 215-71 17 See Tyler Burge (1979)

19 18 L Bonjour

at all to think that the internal structure of my physical and neurophysiological states could somehow by itself determine that I am thinking about weather tather than about the Middle East or the stock market

A functionalist would no doubt say that it is no surprise that Mary could not do this In order to know the complete causal or functional role of my internal states Mary also needs to know about their external causal relations to various things And it might be suggested if Mary knows all of the external causal relations in which my various states stand she will in fact be able ro figure out what I am consciously thinking about at any particular time No doubt the details that pick out any particular object of thought will be very complicated but there is it might be claimed no reason to doubt that in principle she could do this

Here we have a piece of materialist doctrine that again has a status very similar to that of a claim of theology It is obvious that no one ha~ even the beginnings of an idea of how to actually carry out an investigation that would yield a result of this kind-that the only rea~on for thinking that this could be done is the overriding assumption for which we have found no cogent basis that materialism must be true Among a multitude of other difficulties Mary would have to be able to figure out the content of thoughts that are confused or inaccurate or thoughts about imaginary or fictional entities or supernatural entities It is to say the least very hard to see how she could do this on the basis of a knowledge of causal relations to more ordinary sorts of things

But the problem for materialism is in fact even worse than that For as already emphasized it is an undeniable fact about conscious intentional content that I am able for the most part to consciously understand or be aware of what I am thinking about from the inside Clearly I do not in general do this on the basis of external causal knowledge I do not have such knowledge and would not know what to do about it if I did All that I normally have any sort of direct access to if materialism is true is my own internal physical and physiological states and thus my conscious understanding of what I am thinking about at a particular moment must be somehow a feature or result of those internal states alone Causal relations to external things may help to produce the relevant features of the internal states in question but there is no apparent way in which such external relations can somehow be partly constitutive of the fact that my conscious thoughts are about various things in a way of which I can be immediately aware But if these internal states are sufficient to fix the object of my thought in a way that is accessible to my understanding or awareness then knowing about those internal states should be sufficient for Mary a~ well without any knowledge of the external causal relations And yet as we have already seen it seems obvious that this is not the ca~e18

18 Ir is worth noting that the same thing is really true in the ase of qualia as well A persons awareness of one color rather than another when he or she looks at newly mown grass obviously

Against Materialism

Thus we have the basis for an argument that is parallel to Jacksons original argument about qualia Mary knows all the relevant physical facts she is not able on the basis of this knowledge to know what I am consciously thinking about at a particular moment but what I am thinking about at that moment is as surely a fact about the world as anything else therefore complete physical knowledge is not complete knowledge and so materialism is false

One way to further elaborate this point is to consider how it applies to

what is perhaps the most widely held materialist view of intentional content the view popularized by Jerry Fodor and many others that intentional mental states employ an internal language a language of thought19 Fodor calls this view the representational theory of the mind though it might better be called the symbolic theory of the mind For the crucial feature of the view is that the language of thought like any language is composed of symbols items that do not stand for anything by virtue of their intrinsic properties but whose representative character depends instead on the relations in which they stand to other things-for Fodor the sorts of causal relations that are captured in the idea of a causal or functional role1o Just as the word dog could in principle have stood for anything (or nothing at all) and in fact stands for a kind ofanimal rather than something else only because of causal relations that arise from the way it is used so also the symbols in the language of thought stand for whatever they stand for only by virrue of analogous sorts of relations and not in virtue of their intrinsic physical and neurophysiological properties Their intentional character is thus extrinsic not intrinsic

Proponents of the language ofthought rarely have much to say about conscious thoughts of the sort that we are focusing on here But it is clear that on their view what happens when I am consciously thinking about say the Middle East is that in some appropriate location in my overall cognitive operations there occurs a symbol (or set of symbols) that refers to the Middle East This symbol

f like the surrounding context in which it occurs is some neurophysiological state or some constellation of such states No one of course has at present any real

knowledge of the concrete nature of such symbols or their larger contexts but it f I will do no harm to follow Fodor in thinking of a mental blackboard on which

mental symbols are inscribed in appropriate ways Thus for me ro be consciously thinking about the Middle East is for me ro have the mental symbol that refers to

does not depend in a constitutive way on external causal rdations even though it may be causally produced by them Thus in that case too a knowledge of the persons internal physical and neurophysiological states alone should enable Mary to pick out One color rather than the other as the right one But it is even more obvious than in the original ase that this is not so

See eg Jerry Fodor Propositional Attitudes and Methodological Solipsism Considered as a Research Srrategy in Cognitive Science both reprinted in his 198 book Rtpwmll1rions (Cambridge MA MIT Press)

20 See for example Jerry Fodor chapter 4 Fodor has subsequently refined this view in various ways but none that affect issues being raised here

20 L Bonjour Against Materialism 21

the Middle East inscribed in the right way on this blackboard But the symbols reference to the Middle East to repeat depends not on its intrinsic physical or neurophysiological character alone but also on the relations in which it stands to other such symbols and directly or indirectly to the external world

Suppose now that Mary is studying my cognitive operations Suppose that she has somehow isolated what amounts to my mental blackboard and the various symbols written on it Obviously this will not in itself tell her what I am thinking about Even if she could somehow focus onthe specific symbol that refers to the Middle East and tell that it is functioning in a way that determines the object of my conscious thought (even though there is no reason to think that she could in fact do these things) she will not on this basis alone be able to tell what it is that this symbol in fact refers to Nor is there any plausibiliry to the idea that Mary could figure out the reference or meaning of the various mental symbols simply by examining their internal relations (0 each other21 Thus she will need once again to appeal to external causal relations ofvarious sorts

Bur how then am I able to be aware ofor understand from the inside what I am thinking about Once again I have no knowledge of those external relations (and would be very unlikely to be able to figure anything out from them even if I did) All that I plausibly have access to is the mental symbol or symbols and the surrounding system of states and this is apparently nor enough to determine the object of my thought

The only very obvious recourse here for the proponent of a language of thought is to construe my understanding or awareness of what I am thinking about disquotationally in relation to the language of thought When thinking about the Middle East I do so by using some mental symbol And when [ understand or am aware of what I am thinking about it might be suggested I in effect use that very same symbol what I am aware of is that I am thinking about the Middle East (whatever that is-that is whatever that symbol in fact refers (0) If the symbol in question did succeed in referring to the Middle East then this specification of what I am thinking about will refer to the Middle East as well and so will be correct But it is intuitively as obvious as anything could be that my awareness ofwhat I am thinking about normally involves more than this involves actually understanding (at some level of precision detail etc) what the Middle East is in a way that goes beyond merely repeating the same symbol Assuming for the moment that there really is a language of thought I understand my language of thought in a way parallel to the way in which I understand my own public language-and not in the merely disquotational way that could just as well be applied to a language of which I have no understanding at all

Here a proponent of the language of thought may want to reply that the difference in the public language case is merely that one language is a language

2 Such an idea has sometimes at least apparently been suggested For more discussion see 01)

19()8 book In Difmse ofPure Reason pp 174-80

that I successfully use-and that the same is true of my language of thought On this view the intuition that I understand what I am thinking about-or what I am talking about-in any stronger sense one that is not merely disquotational is merely an illusion But here again we have a view that it seems to me would appeal to no one who was not motivated by the conviction that materialism must be true

My conclusion is that the language of thought view has nothing useful to say about the most obvious sort of intentional content the intentional content that is involved in having something explicitly and consciously in mind Nor do I know of any other materialist account that does any better in this regard There is perhaps room for dispute about just how important conscious intentional content is in relation to the causation and explanation of behavior but no plausible way (0 deny that it genuinely exists Thus with respect to intentional content as with the case of qualitative content materialism seems to be utterly bankrupt as a general account of mental states and to be held merely as an article of faith

5 WHAT IS THE ALTERNATIVE

The last two sections serve merely to strengthen and deepen the fundamental objection to materialism already offered in section 2 consciousness genuinely exists materialism can oller no account that explains consciousness (or of the specific varieties thereof) or shows it [0 be merely material in character therefore (at least in the absence of any strong antecedent argument or presumption in favor of materialism) the indicated conclusion is that materialism is false There is more in heaven and earth than is dreamt of in materialist philosophy

But what do I mean by more Here as I see it there is very little that can be said in our current state of knowledge so thar the main result is that we have very little understanding ofconsciousness-or given the arguably central role of consciousness of mentaliry in general

In the first place there is no clear way in which the objections that I have raised against materialism support the classical substance dualist position Positing a separate mental substance that is characterized in almost entirely negative terms does nothing very obvious to explain consciousness in general or qualitative and intentional content in particular As far as I can see the main appeal of substance dualism is that the account of the supposed mental or spiritual substances is fur too vague and sketchy to provide the basis for any very clear argument that such substances could not be the locus of consciousness But this negative point hardly counts as an argument in favor of such a view

The obvious alternative is properry dualism the view that human persons and perhaps other kinds of animals have non-material or non-physical properties in addition to their physical ones with at least the main such properties being

4 5 L Bonjour

materialism arguments by elimination are always dubious in philosophy and never more so than here where the central phenomenon in question (that is consciousness) is arguably something of which we still have little if any real understanding Instead materialism seems to be one of those unfortunate intellectual bandwagons to which philosophy along with many other discipshylines is so susceptible-on a par with logical behaviorism phenomenalism the insistence that all philosophical issues penain [0 language and so many other views that were once widely held and now seem merely foolish Such a comparison is misleading in one important respect however it understates the fervency with which materialist views are often held In this respect matershyialism often more closely resembles a religious conviction-and indeed as I will suggest further in a couple of places below defenses of materialism and especially replies to objections often have a distinctively scholastic or theological flavor

In what follows I will try to substantiate this indictment of materialism by doing the following things First I will look at some of the main considerations that are advanced in favor of materialism in general as opposed to particular materialist views attempting ro show that these are surprisingly insubstanrial and rest mainly on assumptions for which no real defense is offered Second I will look at the overwhelmingly dominant materialist view namely functionalism arguing that it is deeply inadequate in relation ro the problem of consciousness Third I will look at what is widely regarded as the most serious specific problem for materialism in general and functionalism in particular namely the problem of qualitative content or qualia focusing here on a somewhat modified version of Frank Jacksons well-known knowledge argument and trying to show that the objection ro materialism that results is still extremely compelling Fourth I will look at a problem that functionalism is claimed ro handle more successfully the problem of intentional states arguing that there are dear cases of conscious intentional states which materialism in general and functionalism in particular can handle no better than qualia-and for essentially the same reasons Fifth and last I will ask what lessons if any for a more adequate account of conscious mental states can be derived from all of this

1 THE CASE FOR MATERIALISM

One of the oddest things about discussions of materialism is the way in which the conviction that some materialist view must be correct seems to

float free of the defense of any particular materialist view It is very easy to

find people who seem to be saying that while there are admittedly serious problems with all of the specific materialist views it is still reasonable to preshysume that some materialist view must be correct even if we dont know yet

igainst Materialism

which one or that the seeming force of the objections to particular matershyialist views must be balanced against the strength of the underlying case for materialism But why is this supposed to be a reasonable stance to take What arguments or reasons or otherwise compelling intellectual considerations are there that could yield a strong background presumption of this sort in favor of materialism (or create a substantial burden of proof for opponents of materialism)

There are of course arguments against particular versions of dualism mainly against the imeractionist version of Cartesian substance dualism For reasons already mentioned I will set these aside as not constituting in themselves an argument for materialism There is also the inductive generalization from the conspicuous success of materialist science in a wide variety of other areas This undeniably has some modest weight but seems obviously very far from being enough to justifY the strong presumption in question Inductions are always questionable when the conclusion extends to cases that are significantly different from the ones to which the evidence pertains and even most materialists will concede that conscious phenomena are among the most difficult-indeed seemingly the most difficult ofall-for materialist views to handle Thus the fact that materialism has been successful in many other areas does not yield a very strong case that it will succeed in the specific area that we are concerned with

Beyond this there seem to be only rwo related sorts ofgrounds that are offered for a Strongly pro-materialist presumption both of which are quite flimsy when subjected to any real scrutiny

1 L The Principle of Causal Closure

The first and clearer of these rwo grounds appeals to the thesis that the material universe is cl1usI1L~l claud that material things are never causally affected by anything non-material (so that as it is often put physical science can in principle give a completely adequate explanation of any physical occurrence without needing to mention anything non-physical) This thesis is commonly referred [0 as a principle a characterization that leaves its status rather obscure (Philosophers often seem to describe something as a principle when they are inviting their readers to accept it as a basis for further argument even though no clear defense of it has been offered)

I The closure principle does not by itself entail that materialism is true It leaves open both the possibility of non-material realms that are causally isolated from

I the material world and also the possibiliry that epiphenomenalism is true that conscious phenomena are side-effects of material processes that are incapable of having any reciprocal influence on the material world But assuming that the non-material realm in the first possibiliry is supposed to be the locus of conscious phenomena both of these possibilities are extremely unpalatable even paradoxical in essentially the same way The main problem is not as

i

7 6 L Bonjour

is often suggested that they are incompatible with the general common sense intuition that conscious states causally affect bodily behavior A more specific and serious problem is that if either of these possibilities holds then it becomes difficult or seemingly impossible to see how verbal discussions of conscious phenomena-such as this chapter and many others-can be genuinely about them in the way that they seem obviously to be How can people be talking about conscious states or saying anyrhing significant about them if completely adequate causal explanations of their verbal behavior can be given that make no reference to such states Even without invoking any specific version of the causal theory of reference it is hard to see why verbal discussions that are entirely unaffected by what they purport to be about should be taken seriously Thus while a number of philosophers have in recent times been seemingly tempted by epiphenomenalism it appears that they can have been genuinely advocating such a view about conscious states only if the view itself is false 3

For these reasons the argument from the principle of causal closure to the truth of materialism is quite strong even if not fully conclusive But why is the principle of causal closure itself supposed to be so obviously correct Clearly this principle is not and could not be an empirical result no empirical investigation that is at all feasible (practically or morally) could ever establish that human bodies the most likely locus of such external influence are in fact never affected even in small and subtle ways by non-material causes Weare told that scientists accept this principle and often that most philosophers accept it as well But do they have any compelling reasons for such acceptance Or is this vaunted principle nothing more than an unargued and undefended assumption-a kind of intellectual prejudice in the literal meaning of the word

Taken in the abstract apart from any appeal to a specific account of conscious mental phenomena I have no idea whether the principle of causal closure is true or not More importantly I cannot imagine how to rationally decide whether it is true without first arriving at a defensible account of conscious mental states It seems utterly obvious that mental states do causally affect the material realm probably by causally affecting the actions of human bodies in general but (as just argued) at least more narrowly by causally affecting verbal discussions of these matters If a materialist account of conscious states is correct then the principle of causal closure seems likely to be true But if no such account is correct then the principle is almost certainly false Thus to argue for the truth of materialism or for a strong presumption in favor of materialism by appeal to the principle of causal closure is putting the cart in quite a flagrant way before the horse

3 This problem seems ro be the main reason for Jacksons abandonment of his previous antishymaterialist stance (Jackson never rook seriously the possibility that the non-material qualia for which he was arguing might causally affect the material world)

Against Materialism

12 The Appeal to Naturalism

A second sort of defense of a general presumption in favor of materialism appeals to the general idea of naturalism Here again we have a view like materialism itself to which many many philosophers pay allegiance while offering little by way of clear argument or defense but here the view itself is much harder to

pin down in a precise way Indeed even more striking than the absence of any very clear arguments is the fact that many recent philosophers seem so eager to commit themselves to naturalism-to fly the naturalist flag as it were-while showing little agreement as to what exactly such a commitment involves Thus naturalism seems to be even more obviously an intellectual bandwagon than materialism (In addition naturalism for some of those who use the term seems to just amount to materialism which would make an argument from naturalism to materialism entirely question-begging)

Is there any genuine support for a materialist presumption to be found in the viciniry of naturalism One version of naturalism is the idea that metaphysical issues-or philosophical issues generally-should be dealt with through the use of the methods ofnatural science If this is accepted and ifit is true that following the methods of natural science leads plausibly to an endorsement of materialism then at least some presumption in favor of materialism might follow But both of the needed suppositions are in fact extremely dubious to say the least There is simply no good reason to think that the methods of natural science exhaust the methods of reasonable inquiry-indeed as has often been pointed out there is no plausible way in which that claim itself can be arrived at using those methods Nor is there any very clear reason to think that applying the methods of natural science to the question of whether materialism is true assuming that one could figure out some reasonably clear way to do that would lead to the conclusion that materialism is correct Such a conclusion is obviously not within the purview of physics but it is also not within the purview of psychology especially as currently practiced As was true with closure there is no doubt that many (but not all) natural scientists assume the truth of materialism but the question is whether they have any good reason for such an assumption-a reason that would itself have to transcend their strictly scientific claims and competence 4

Thus while the murkiness of the discussions of naturalism makes it harder to be sure naturalism like closure does not seem to yield an independently defensible presumption in favor of the truth of materialism If there is any better

bull Lurking here is the difficult issue of what sons of entities or properties count as material or physical Is there any good war ro delimit the realm of the material that does not preclude further discoveries in physics but also does not trivialize the categoty by allowing it to include anything that people in departments labeled Physics might eventually come to study This IS anything but a trivial problem but I have no space here to pursue it further

8 9 L Bonjour

rea~n or basis for such a presumption that is prior to and independent of the defense of some particular materialist view I have no idea what it might be

2 FUNCTIONALISM AND CONSCIOUSNESS

The upshot of the previous section is that the case for materialism must rest almost entirely on the defense of particular materialist views and not to any substantial extent on any background presumption So what materialist views are there The answer I think is that once both logical behaviorism and various versions of e1iminativism are set aside as toO implausible to be taken seriously-something that I will assume here without any further discussion-there is only one main materialist view namely functionalism with no very serious prospect that any others will emerge And the fundamental problem for materialism I will suggest is that functionalism offers no account at all of consciousness and seems in principle unable to do so

What gives rise to the mind-body problem in the first place and poses the essential problem that any adequate version of materialism mUSt solve is the fact that conscious mental states as we are aware of them do not present a material appearance-do not seem as we experience them to be material in their makeup in any apparent way Thus a view which holds that everything that exists is material must either (a) deny the very existence of such states as e1iminativism does or else (b) explain how states and correlative properties that do not initially seem to be material in nature can nevertheless turn out to be so A view that takes the latter alternative must give an account of the nature of such states and properties that both accurately reflects their character as experienced and explains how they can nonetheless be entirely material in their makeup And this I suggest is something that has never been successfully done

The starting point for modern versions of materialism was the central-state identity theory particularly the version advocated in a famous paper by J J c Smart (1959) Smart recognized that the truth of materialism can only be an empirical discovery not something knowable a priori For this to be so he argued the various mental states in question must be conceived in a topicshyneutral way a way that makes it possible for them to be merely material in character without implausibly requiring that this be so Only in relation to such a conception would it be possible to discover empirically that such a state is in fact a neurophysiological state of some kind

But for this to work it is crucially important that the topic-neutral conshyception in question be adequate to capture the essential features of mental

Against Materialism

states-something about which Smart was less clear than he might have been For only if this is so will it be the case that showing that the conception offered can be realized by a material state can establish that mental states might in fact be merely material states thereby allowing the rest of the argument to proceed on grounds of simplicity as Smart suggests Whereas if the proposed topic-neutral conception leaves out essential features of mental states-such as consciousness-then the mct that material states can satisfy that conception will be insufficient to explain how mental states might just be material states (Smarts own attempt at a topic-neutral characterization fails to distinguish conscious mental states from whatever else might be going on in the person under a particular set of circumstances)

As in Smarts view functionalism in effect attempts to offer a topic-neutral characterization ofmental states one which will allow but not require that they be essentially material in characters The more general functionalist characterization is in terms of causal role a mental state is characterized by its causal relations to sensory inputs behavioral outputs and other mental states of the same sort The functionalist then proceeds to argue that the states thus characterized could perfectly well be material srates even though the functional characterization does not require rhis A further widely discussed aspect of the view is that different material states could satisfy the functionalist characterization of a particular mental state in different SOrtS ofcreatures or even in the same creature at different times so that (on the most standard version) a material state realizes a functionally characterized state but is not strictly identical with it

But the deepest problem for the functionalist is that the characterization of mental states in terms of causal role say~ nothing at all about consciousness or conscious character There is no apparent reason why a stare thar realizes a particular causal role would thereby need to have any specific sort of conscious character (the point made by the familiar reversed spectrum cases)-or indeed any conscious character at all Thus to point out that a physical state could realize such a causal role really does nothing at all to explain how a conscious state could be (or be realized by) a merely physical state In this way functionalism fails utterly to offer any explanation or account of the most important and conspicuous feature of mental states-or at the very least of a very important and conspicuous feature

It may seem hard to believe that a view that has been held by so many people for so long can be so easily shown to be inadequate in a fundamental way but I think that this is nonetheless so The only solution would be

5 This way of looking at functionalism is explicit in David Lewiss discussion of one of the earliest versions of the view in his 1966 paper An Argument for the Identity Theory Journal of Philosophy 63 17-25 see p 20

10 L Bonjour Against Materialism 11

to offer some supplementary account of what material features give rise to

conscious experience But I know of no such account at least none with any real plausibiliry6

This difficulty with materialism in general and functionalism in particular has of course occasionally been recognized But it still seems to have had remarkably little impact on the widespread acceptance of materialist views I have no very good explanation to offer of this though part of the reason is perhaps the prevailing tendency to approach the philosophy of mind from a third-person neo-behaviorist perspective in which consciousness is largely or entirely ignored (But on this issue it is hard to distinguish the chickens from the eggs)

My basic case against materialism is complete at this point there is no good reason for any strong presumption in favor of materialism and the main materialist view fails to offer any real explanation of a central aspect of mental states namely their conscious character meaning that there is no good reason to think that it is correct as an account of such states But though this very simple argument seems to me entirely compelling I will elaborate it further in the next two sections by focusing on the two main specific kinds of mental states The version of the argument that applies to states with qualitative content is very familiar even though I think that its full force has still not been generally appreciated In contrast the application of essentially the same basic argument to conscious states with intentional content has received far less attention

3 THE PROBLEM OF QUALITATIVE CONSCIOUSNESS

MARY REDUX

Though functionalism fails to adequately account for consciousness of any SOrt

perhaps the most conspicuous aspect ofthis failure penains to qualitative content the sort of content involved in experiences of color and sound and of like pains and itches This point has been made in many ways but the most straightforward and compelling in my view is still the so-called argument initially suggested by Thomas Nagel in relation to the experiences of bats and later developed by Frank Jackson using his famous example of black-and-white Mary on which] will mainly focus here (As most will know

6 One possibiliry is the so-allIed higher order thought theory which holds that consciousne arises when one mental state is the object of a second higher order mental state I have no room here to consider this view in detaiL But the basic-and obvious-problem with such a view is that there is no reason why there could not be a hierarchy of sort even one with many more levels in which there was no consciousness involved at alL (For some elaboration see my contribution to

Bonjour and Sos (2003) pp 65-8) 7 See for example Colin McGinn (1989) David Chalmers (1995) and David Chalmers (1996)

Jackson has changed his mind about this argument and now rejects it though his reasons seem to me unpersuasive)

I will assume here that Jacksons original version of the saga of Mary is familiar enough to require only a brief summation Mary is a brilliant neurophysiologist who lives her entire life acquires her education and does all of her scientific work in a black-and-white environment using black-and-white books and blackshyand-white television for all of her learning and research In this way we may suppose she comes to have a complete knowledge of all the physical facts in neurophysiology and related fields together with their deductive consequences insofar as these are relevanr-thus arriving at as complete an understanding of human functioning as those sciences can provide In particular Mary knows the functional roles of all of the various neurophysiological states including those pertaining to visual perception by knowing their causal relations to sensory inputs behavioral outputs and other such states But despite all of this knowledge Mary apparently does nor know all that there is to know about human mental states for when she is released from her black-and-white environmenr and allowed to view the world normally she will by viewing objects like ripe tomatoes learn what it is like to see something red and analogous things about other qualitative experiences But then comments Jackson it is inescapable that her previous knowledge was incomplete But she had all the physical information Ergo there is more to have than that and Physicalism is falses

Despite the initial force of this rather simple argumenr materialists have nOt been persuaded and the literature comprising materialist responses to the Mary example is very large 9 One thing to say about these responses is that few if any of them are even claimed ro have any substanrial independent plausibility instead they are PUt forward in a way that takes for granted the sort of general presumption in favor of materialism and correlative burden of proof for antishymaterialist views that J have argued does not genuinely exist A full discussion of these responses is impossible here but there are some main themes that can be

dealt with in a general way 10 One of these is the suggestion that although something new when she leaves the black-and-white

room what she acquires is not a knowledge of a new foct (or facts) bur rather something else A second is the suggestion that what she does something like a new ability perhaps more specificall representational ability And if these rwo themes are

bull Frank Jackson (1982) p 130 see also Frank Jackson (1986) bull Many of these discussions are collected in Ludlow Nagasawa and

)0 For a useful taxonomy of the various possible materialist responses see Robert Van Gulick So Many Ways of Saying No to Mary in Ludlow Nagasawa and Stoljar (2004) pp 365-405 (This is of the places where the materialist discussion bears a striking similariry to scholastic theology one can easily imagine a complacent theist writing an artide entitled So Many Ways to Answer the Problem of Evil)

12 L Bonjour i Against Materialism 13

fresult is that there is nothing about the Mary example that is incompatible with Though we have made things vastly easier for Mary by focusing on twO cases the truth of materialism involving colors with which she is now familiar rather than asking her to figure

Particularly in light of rhe general materialist failure to provide an account out on the basis ofher overall physicalneurophysiologicalfunctional knowledge I of conscious experience I doubt very much whether any response of this son would seem even mildly convincing to anyone who was not already determined to adhere to materialism come what may But the first of these fWO themes does at least point to a kind of lacuna in Jacksons original account of the case if Mary learns new facts what exactly are they Indeed in his response to an early version of this suggestion Jackson is reduced to invoking the problem of other minds as a (very) indirect basis for thinking that factual knowledge of some son is involved II

It is however surprisingly easy to modifY the original case in a way that makes it utterly dear that there are facts that Mary does not know while she is in the black-and-white room and will learn when she emerges Suppose that while she is still in the otherwise black-and-white environment fWO color samples are brought in one a sample of a fairly bright green approximately the color of newly mown grass and the other a sample of a fairly bright red approximately the color of a fire engine Mary is allowed to view these samples and even to know that they are fWO of the colors that she has learned about in her black-and-white education She is not however told the standard names of these colors nor is she allowed to monitor her own neurophysiology as she views them

We now remind Mary of two specific ca~es that she has studied morougIllV and about which she knows all the physic~l One of these is a case where a person was second is a case where a person was lookmg at a tell Mary that one of these people had an experience predominar one of the colors with which she is now familiar and that the other person had an experience predominantly involving the other color but of course not which was which If we call the colors presented by the samples color A and color B Mary now knows that one of the two following pairs of claiflls is true

(l) The experience of freshly mown grass predominantly involves color A and the experience of a newly painted fire engine predominantly involve~ color B

(2) The experience of freshly mown grass predominantly involves color B and the experience ofa newly painted fire engine predominantly involves color A

But can she tell on the basis of her black-and-white knowledge together with her new familiarity with the two colors whether it is (1) or (2) which is true (Notice carefully that there is no apparent problem with her understanding of these claims)

l See Jackson (1986 294)

what color experiences in general are like it still seems quite dear for essentially the same reasons that were operative in the original case that she will have no more success with her much more limired task Just as there was nothing in the physical account that could tell her what an experience of red was like so there is still nothing in the physical account of the fire engine case that could definitively pick out the color of one of the samples as opposed to the other 12

And yet whichever of (1) and (2) is true states a foct (or facts) in as robust a sense as one could want-a fact that Mary will learn when she emerges from the black-and-white room and is allowed to view ordinary objects of various kinds

Moreover if there are abilities that result from experiencing the two colors in question Mary presumably can acquire them on the basis of the samples Consider for example Harmans suggestion l3 that what Mary acquires in the original case when she leaves the black-and-white room and sees red for the first time is a perceptual concept of red one that essentially involves being disposed to form perceptual representations involving it in the presence ofcausal stimulation of the right SOrt-sO that she cannot acquire that concept in the original version of the black-and-white room There is much that is questionable about the idea ofsuch a concept but if there is indeed such a thing then Mary in the new version of the case presumably can acquire it by viewing the red sample

more than one sample is for some reason required but it would be easy the new version of the case to allow for that) So we may

suppose Mary has the perceptual concept of red and the perceptual concept of green but she still cannot figure our from her physical knowledge which of these concepts is being employed by the people in the cases she has studied Yet this too is a fact and ifmaterialism isrrue an entirely physical fact So why cant she know it

Here as far as I can see there are only fWO possible moves for the materialist which are even marginally worth considering One is the suggestion that Mary already knows the facts in question as a part of her overall physical knowledge but that she knows them under a different guise or mode of presentation than that under which she will come to know them when she leaves the blackshyand-white room This idea can be developed in different ways and with enormous technical ingenuity But does it really have any serious plausibility Imagine that

J2 As Jackson emphasizes in Jackson (1986 295) it is nO enough for Mary (0 be able to conjecture or guess at the answerO this question For physicalism to be true the fact in question must actually be contained in her physical knowledge

U Gilbert Harman (1990) The Intrinsic Quality of Experience Philosophical Pmpectives 4 3J~52 at pp 44~5 Harman does not actually melllion the Maty case as sucn focusing instead on a person who is blind from birth but still learns all the physical and functional facts of color perception But he does cite Jackson (along with Nagel) as the source of tne objection he is discussing

14 15 L Bonjour

Mary in our modified version of the case having finally experienced real colors is eager to find out more about these intriguing features of the world about which she has been kept in ignorance She wants very much for example to know whether it is (I) or (2) that captures the relevant facts about cases of that sort-and is seriously frustrated about being kept in ignorance any longer Suppose that we respond to her frustration by informing her that she already knows the very facts that she is so eager to learn Surely she would not be satisfied How might she respond

I think we can imagine Mary saying something like this

You philosophers are really amazing The idea that I ar(atiy know the facts I am interested in-indeed all facts of that general kind-is simply preposterous I know all of the physical details but none of them tells me which of the properties I have just experienced on the basis of the samples is realized in each of the two cases If you suggest to me that there arent really novel properties but rather novel concepts or ways of representing or whatever then (while finding that suggestion itself pretry hard to swallow) I would still insist that which concept or way of representing is involved in each case is still something that my physical knowledge doesnt give me any clue about Perhaps as you say there is some clever or complicated way in which the things I want to know are related to the physical things I do know-maybe there is even some metaphysically necessary connection between them (assuming that it is kosher for materialists to believe in such things) Anything like that however just adds to the list of facts that my physical knowledge doesnt reveal to me I am a scientist and not a philosopher so Im not really sure which conception of a fact is the right one (All of the ones you suggest seem pretty weird) But there is undeniably something that I want to know-something that is true about the world-that cant be learned on the basis of all my physical knowledge And that means that the physical story isnt in fact the whole story

Not surprisingly I think that the response I have imagined for Mary is exactly right-that any way of understanding or individuating facts according to which some piece of Marys physical knowledge and either (1) or (2) above turn out to be formulations of the same fact is a conception of fact that is simply too intuitively implausible to be taken seriously

The other possible materialist response is to gram that Mary will learn new facts when she emerges from the black-and-white room in the modified version of the case but to insist that these are nonetheless still physical facts On this view what the case shows is that it is impossible for Mary to acquire complete physical knowledge in the black-and-white room One way to put it is to say that while she can learn all the objective physical facts there are still certain subjective physical facts l4 that she cant learn One can learn what it feels like subjectively to be an organism ofsuch-and-such a general physical description in such-and-such a specific physical state only by actually realizing that condition But that it feels

14 See Van Gulick (2004) for one version of this suggestion

Agaimt Materialism

a certain way or involves a certain sort of conscious experience is still on this view an entirely physical fact

I have to admit that I find it nearly impossible to take this response seriously The only argument for it seems to be an appeal to a background presumption in favor of materialism that is so strong as to make it allegedly reasonable to claim that any fact there is must be a material fact even ifwe cant see in any clear way how it could be a material fact Materialism as we have already seen offers no real account or explanation of consciousness and so also no reason for thinking that there is any subjective experience at all involved in being in a certain material state Thus to advance a view of this sort is in effect just to insist that no fact of any sort can be allowed to refute materialism and thus that any possibiliry of this sort must simply be absorbed into the materialist view however inexplicable in materialist terms it may be (It is not much of a stretch to imagine the materialist saying that we must first believe in order than we may understand)

Thus the modified version of the Mary case seems to present an objection to materialism in general (and functionalism in particular) that is about as conclusive as philosophical arguments ever get However exactly they should be characterized there are facts that Mary cannot know on the basis ofher complete physicalneurophysiologicalfunctional knowledge even when she is given the sort of limited experience needed to understand the claims in question and to acquire any abilities that might be relevant These facts do not seem to be material facts and there is no basis that is not utterly arbitrary and question-begging for supposing that they are Thus we have the strongest of reasons for holding that the materialist account of realiry is incomplete and hence that materialism is

false

4 THE PROBLEM OF CONSCIOUS INTENTIONAL

CONTENT

Qualia of the son involved in the Mary case are widely recognized to pose a serious problem for materialist views and it is not too hard to discern occasional misgivings in this respect under the faltade of materialist confidence But as already mentioned above it is widely assumed that materialism is in much better shape with regard to intentional mental states propositional attitudes and other states that involve aboutness I believe however that this is almost entirely an illusion-that the problems for materialism are just as serious in this area with consciousness being once again the central focus

Materialist accounts of intentional states tend to focus mainly on dispositional states such as beliefs and desires Given the central role of such states in explanations of behavior this is in some ways reasonable enough But such a focus tends to neglect or even ignore the existence of comcious intentional states-even though having conscious thoughts that P is surely one of the

16 L Bonjour

central things that having a dispositional belief that P disposes one to do A focus on belief in particular also has the unfortunate effect of making externalist accounts of intentional content seem more plausible than they possibly could if the emphasis were on conscious intentional states

For these reasons I will focus here explicitly on conscious thoughts As I sit writing this chapter a variety of conscious thoughts pass through my mind Many of these involve the assertion or endorsement of various propositions that materialism cannot account for consciousness that the trees outside my window are very bare that the weather looks cold and dank that the situation in the Middle East looks grim and so on Other thoughts are also propositional but in a way that does not involve assertion my conscious desire to get several pages written before lunch the hope that the stock market will continue to rise and so on It is doubtful that conscious thought must always be propositional in character but it will in any case simplify the issues to be discussed if we largely ignore the propositional aspect of these various thoughts and focus simply on their being conscious thoughts of or about various things or kinds of things materialism the trees the Middle East the stock market and so on

One crucial feature of such conscious thoughts is that when I have them I am in general consciously aware of or consciously understand or grasp what it is that I am thinking about (and also what I am thinking about it) When I think that the trees outside my window are bare I consciously understand that it is certain trees that I am thinking about (and along with this what sort of thing a tree is and which trees I have in mind) What exactly this conscious grasp of the object of thought involves varies from case to case and is sometimes not easy to precisely specify Moreover as will emerge it is something of which I think we presently have no real explanatory account of any substance But its existence is I submit completely undeniable Indeed being able in this way to consciously think about things to have them in mind is in many ways the most central and obvious feature of our mental lives

It is obvious that a persons conscious grasp of the object of their thought of what they are thinking about can vary on a number of dimensions it may be more or less precise more or less detailed more or less clear more or less complete But contrary to what is sometimes suggested it is rarely if ever merely disquotational in character Perhaps (though I doubt it) there are cases where a scientifically untutored person is thinking about eg electrons and where their sole grasp of what they are thinking is that it is what is referred to in their society or community by the word electron -so that what they are thinking about is in effect electrons (whatever they are) But this is surely not the ordinary situation when we think about various things 15

15 Notice that even a thought about what the relevant Ioci~tal ocp~rts mean by electrons would have to involve a non-disquotational element in the reference to those experts and also in the reference to the word to think about whatever it is that the societal experts mean by electrons

Against Materialism 17

Moreover the existence of conscious intentional content is perfectly compatshyible with the existence of an externalist dimension of thought content-though not with the view that all content is external If as in Putnams famous example a person is thinking about earthly water at a time prior to the discovery of its chemical composition there is no reason to deny that they are in a sense thinking about H 20 But in such a situation the aspect of being about H 20 will obviously not be part of their conscious internal grasp of what they are thinking about in the way that the more superficial aspects of water will 16 And in a somewhat parallel way the person in Burges famous example who thinks he has arthritis in his thigh and to whom our standard belief-ascription practices ascribe a belief about arthritis (where this is among other things a disease that only occurs in joints) obviously does not consciously grasp the disease that he is thinking about in a way that involves this specific feature of itY But it is nonetheless impossible to describe either example in a convincing way without presupposing that the people in question do have something consciously in mind a substance having the superficial properties of water in Putnams example and a disease having certain fairly specific features in Burges Thus while it is possible to dispute the relative importance ofconscious internal thought content and external thought content of which the subject is not conscious examples of this sort provide no basis at all for denying that conscious internal content exists

The issue I want to raise here is whether a materialist view can account for the sort of conscious intentional content just characterized Can it account for conscious thoughts being about various things in a way that can be grasped or understood by the person in question In a way the answer has already been given Since materialist views really take no account at all of consciousness they obviously offer no account of this particular aspect of it But investigating this narrower aspect of the issue can still help to deepen the basic objection to

materialism Here it will be useful to bring the brilliant neurophysiologist Mary briefly

back onto the scene even though the black-and-white aspect of her situation is no longer relevant Suppose that Mary studies me as a subject and comes to

have a complete knowledge of my physical and neurophysiological makeup as I am thinking these various thoughts Can she determine on that basis what I am consciously thinking about at a particular moment

One thing that seems utterly clear is that she could not do this merely on the basis of knowing my internal physical characteristics-as it is sometimes put

I knowing everything physical that happens inside my skin There is no reason t f

is not the same thing as thinking about whatever is meant by electrons by whatever is meant by the societal experts

I 16 See Hilary Putnam (1975a) (Cambridge Cambridge Universiry Press) pp 215-71 17 See Tyler Burge (1979)

19 18 L Bonjour

at all to think that the internal structure of my physical and neurophysiological states could somehow by itself determine that I am thinking about weather tather than about the Middle East or the stock market

A functionalist would no doubt say that it is no surprise that Mary could not do this In order to know the complete causal or functional role of my internal states Mary also needs to know about their external causal relations to various things And it might be suggested if Mary knows all of the external causal relations in which my various states stand she will in fact be able ro figure out what I am consciously thinking about at any particular time No doubt the details that pick out any particular object of thought will be very complicated but there is it might be claimed no reason to doubt that in principle she could do this

Here we have a piece of materialist doctrine that again has a status very similar to that of a claim of theology It is obvious that no one ha~ even the beginnings of an idea of how to actually carry out an investigation that would yield a result of this kind-that the only rea~on for thinking that this could be done is the overriding assumption for which we have found no cogent basis that materialism must be true Among a multitude of other difficulties Mary would have to be able to figure out the content of thoughts that are confused or inaccurate or thoughts about imaginary or fictional entities or supernatural entities It is to say the least very hard to see how she could do this on the basis of a knowledge of causal relations to more ordinary sorts of things

But the problem for materialism is in fact even worse than that For as already emphasized it is an undeniable fact about conscious intentional content that I am able for the most part to consciously understand or be aware of what I am thinking about from the inside Clearly I do not in general do this on the basis of external causal knowledge I do not have such knowledge and would not know what to do about it if I did All that I normally have any sort of direct access to if materialism is true is my own internal physical and physiological states and thus my conscious understanding of what I am thinking about at a particular moment must be somehow a feature or result of those internal states alone Causal relations to external things may help to produce the relevant features of the internal states in question but there is no apparent way in which such external relations can somehow be partly constitutive of the fact that my conscious thoughts are about various things in a way of which I can be immediately aware But if these internal states are sufficient to fix the object of my thought in a way that is accessible to my understanding or awareness then knowing about those internal states should be sufficient for Mary a~ well without any knowledge of the external causal relations And yet as we have already seen it seems obvious that this is not the ca~e18

18 Ir is worth noting that the same thing is really true in the ase of qualia as well A persons awareness of one color rather than another when he or she looks at newly mown grass obviously

Against Materialism

Thus we have the basis for an argument that is parallel to Jacksons original argument about qualia Mary knows all the relevant physical facts she is not able on the basis of this knowledge to know what I am consciously thinking about at a particular moment but what I am thinking about at that moment is as surely a fact about the world as anything else therefore complete physical knowledge is not complete knowledge and so materialism is false

One way to further elaborate this point is to consider how it applies to

what is perhaps the most widely held materialist view of intentional content the view popularized by Jerry Fodor and many others that intentional mental states employ an internal language a language of thought19 Fodor calls this view the representational theory of the mind though it might better be called the symbolic theory of the mind For the crucial feature of the view is that the language of thought like any language is composed of symbols items that do not stand for anything by virtue of their intrinsic properties but whose representative character depends instead on the relations in which they stand to other things-for Fodor the sorts of causal relations that are captured in the idea of a causal or functional role1o Just as the word dog could in principle have stood for anything (or nothing at all) and in fact stands for a kind ofanimal rather than something else only because of causal relations that arise from the way it is used so also the symbols in the language of thought stand for whatever they stand for only by virrue of analogous sorts of relations and not in virtue of their intrinsic physical and neurophysiological properties Their intentional character is thus extrinsic not intrinsic

Proponents of the language ofthought rarely have much to say about conscious thoughts of the sort that we are focusing on here But it is clear that on their view what happens when I am consciously thinking about say the Middle East is that in some appropriate location in my overall cognitive operations there occurs a symbol (or set of symbols) that refers to the Middle East This symbol

f like the surrounding context in which it occurs is some neurophysiological state or some constellation of such states No one of course has at present any real

knowledge of the concrete nature of such symbols or their larger contexts but it f I will do no harm to follow Fodor in thinking of a mental blackboard on which

mental symbols are inscribed in appropriate ways Thus for me ro be consciously thinking about the Middle East is for me ro have the mental symbol that refers to

does not depend in a constitutive way on external causal rdations even though it may be causally produced by them Thus in that case too a knowledge of the persons internal physical and neurophysiological states alone should enable Mary to pick out One color rather than the other as the right one But it is even more obvious than in the original ase that this is not so

See eg Jerry Fodor Propositional Attitudes and Methodological Solipsism Considered as a Research Srrategy in Cognitive Science both reprinted in his 198 book Rtpwmll1rions (Cambridge MA MIT Press)

20 See for example Jerry Fodor chapter 4 Fodor has subsequently refined this view in various ways but none that affect issues being raised here

20 L Bonjour Against Materialism 21

the Middle East inscribed in the right way on this blackboard But the symbols reference to the Middle East to repeat depends not on its intrinsic physical or neurophysiological character alone but also on the relations in which it stands to other such symbols and directly or indirectly to the external world

Suppose now that Mary is studying my cognitive operations Suppose that she has somehow isolated what amounts to my mental blackboard and the various symbols written on it Obviously this will not in itself tell her what I am thinking about Even if she could somehow focus onthe specific symbol that refers to the Middle East and tell that it is functioning in a way that determines the object of my conscious thought (even though there is no reason to think that she could in fact do these things) she will not on this basis alone be able to tell what it is that this symbol in fact refers to Nor is there any plausibiliry to the idea that Mary could figure out the reference or meaning of the various mental symbols simply by examining their internal relations (0 each other21 Thus she will need once again to appeal to external causal relations ofvarious sorts

Bur how then am I able to be aware ofor understand from the inside what I am thinking about Once again I have no knowledge of those external relations (and would be very unlikely to be able to figure anything out from them even if I did) All that I plausibly have access to is the mental symbol or symbols and the surrounding system of states and this is apparently nor enough to determine the object of my thought

The only very obvious recourse here for the proponent of a language of thought is to construe my understanding or awareness of what I am thinking about disquotationally in relation to the language of thought When thinking about the Middle East I do so by using some mental symbol And when [ understand or am aware of what I am thinking about it might be suggested I in effect use that very same symbol what I am aware of is that I am thinking about the Middle East (whatever that is-that is whatever that symbol in fact refers (0) If the symbol in question did succeed in referring to the Middle East then this specification of what I am thinking about will refer to the Middle East as well and so will be correct But it is intuitively as obvious as anything could be that my awareness ofwhat I am thinking about normally involves more than this involves actually understanding (at some level of precision detail etc) what the Middle East is in a way that goes beyond merely repeating the same symbol Assuming for the moment that there really is a language of thought I understand my language of thought in a way parallel to the way in which I understand my own public language-and not in the merely disquotational way that could just as well be applied to a language of which I have no understanding at all

Here a proponent of the language of thought may want to reply that the difference in the public language case is merely that one language is a language

2 Such an idea has sometimes at least apparently been suggested For more discussion see 01)

19()8 book In Difmse ofPure Reason pp 174-80

that I successfully use-and that the same is true of my language of thought On this view the intuition that I understand what I am thinking about-or what I am talking about-in any stronger sense one that is not merely disquotational is merely an illusion But here again we have a view that it seems to me would appeal to no one who was not motivated by the conviction that materialism must be true

My conclusion is that the language of thought view has nothing useful to say about the most obvious sort of intentional content the intentional content that is involved in having something explicitly and consciously in mind Nor do I know of any other materialist account that does any better in this regard There is perhaps room for dispute about just how important conscious intentional content is in relation to the causation and explanation of behavior but no plausible way (0 deny that it genuinely exists Thus with respect to intentional content as with the case of qualitative content materialism seems to be utterly bankrupt as a general account of mental states and to be held merely as an article of faith

5 WHAT IS THE ALTERNATIVE

The last two sections serve merely to strengthen and deepen the fundamental objection to materialism already offered in section 2 consciousness genuinely exists materialism can oller no account that explains consciousness (or of the specific varieties thereof) or shows it [0 be merely material in character therefore (at least in the absence of any strong antecedent argument or presumption in favor of materialism) the indicated conclusion is that materialism is false There is more in heaven and earth than is dreamt of in materialist philosophy

But what do I mean by more Here as I see it there is very little that can be said in our current state of knowledge so thar the main result is that we have very little understanding ofconsciousness-or given the arguably central role of consciousness of mentaliry in general

In the first place there is no clear way in which the objections that I have raised against materialism support the classical substance dualist position Positing a separate mental substance that is characterized in almost entirely negative terms does nothing very obvious to explain consciousness in general or qualitative and intentional content in particular As far as I can see the main appeal of substance dualism is that the account of the supposed mental or spiritual substances is fur too vague and sketchy to provide the basis for any very clear argument that such substances could not be the locus of consciousness But this negative point hardly counts as an argument in favor of such a view

The obvious alternative is properry dualism the view that human persons and perhaps other kinds of animals have non-material or non-physical properties in addition to their physical ones with at least the main such properties being

7 6 L Bonjour

is often suggested that they are incompatible with the general common sense intuition that conscious states causally affect bodily behavior A more specific and serious problem is that if either of these possibilities holds then it becomes difficult or seemingly impossible to see how verbal discussions of conscious phenomena-such as this chapter and many others-can be genuinely about them in the way that they seem obviously to be How can people be talking about conscious states or saying anyrhing significant about them if completely adequate causal explanations of their verbal behavior can be given that make no reference to such states Even without invoking any specific version of the causal theory of reference it is hard to see why verbal discussions that are entirely unaffected by what they purport to be about should be taken seriously Thus while a number of philosophers have in recent times been seemingly tempted by epiphenomenalism it appears that they can have been genuinely advocating such a view about conscious states only if the view itself is false 3

For these reasons the argument from the principle of causal closure to the truth of materialism is quite strong even if not fully conclusive But why is the principle of causal closure itself supposed to be so obviously correct Clearly this principle is not and could not be an empirical result no empirical investigation that is at all feasible (practically or morally) could ever establish that human bodies the most likely locus of such external influence are in fact never affected even in small and subtle ways by non-material causes Weare told that scientists accept this principle and often that most philosophers accept it as well But do they have any compelling reasons for such acceptance Or is this vaunted principle nothing more than an unargued and undefended assumption-a kind of intellectual prejudice in the literal meaning of the word

Taken in the abstract apart from any appeal to a specific account of conscious mental phenomena I have no idea whether the principle of causal closure is true or not More importantly I cannot imagine how to rationally decide whether it is true without first arriving at a defensible account of conscious mental states It seems utterly obvious that mental states do causally affect the material realm probably by causally affecting the actions of human bodies in general but (as just argued) at least more narrowly by causally affecting verbal discussions of these matters If a materialist account of conscious states is correct then the principle of causal closure seems likely to be true But if no such account is correct then the principle is almost certainly false Thus to argue for the truth of materialism or for a strong presumption in favor of materialism by appeal to the principle of causal closure is putting the cart in quite a flagrant way before the horse

3 This problem seems ro be the main reason for Jacksons abandonment of his previous antishymaterialist stance (Jackson never rook seriously the possibility that the non-material qualia for which he was arguing might causally affect the material world)

Against Materialism

12 The Appeal to Naturalism

A second sort of defense of a general presumption in favor of materialism appeals to the general idea of naturalism Here again we have a view like materialism itself to which many many philosophers pay allegiance while offering little by way of clear argument or defense but here the view itself is much harder to

pin down in a precise way Indeed even more striking than the absence of any very clear arguments is the fact that many recent philosophers seem so eager to commit themselves to naturalism-to fly the naturalist flag as it were-while showing little agreement as to what exactly such a commitment involves Thus naturalism seems to be even more obviously an intellectual bandwagon than materialism (In addition naturalism for some of those who use the term seems to just amount to materialism which would make an argument from naturalism to materialism entirely question-begging)

Is there any genuine support for a materialist presumption to be found in the viciniry of naturalism One version of naturalism is the idea that metaphysical issues-or philosophical issues generally-should be dealt with through the use of the methods ofnatural science If this is accepted and ifit is true that following the methods of natural science leads plausibly to an endorsement of materialism then at least some presumption in favor of materialism might follow But both of the needed suppositions are in fact extremely dubious to say the least There is simply no good reason to think that the methods of natural science exhaust the methods of reasonable inquiry-indeed as has often been pointed out there is no plausible way in which that claim itself can be arrived at using those methods Nor is there any very clear reason to think that applying the methods of natural science to the question of whether materialism is true assuming that one could figure out some reasonably clear way to do that would lead to the conclusion that materialism is correct Such a conclusion is obviously not within the purview of physics but it is also not within the purview of psychology especially as currently practiced As was true with closure there is no doubt that many (but not all) natural scientists assume the truth of materialism but the question is whether they have any good reason for such an assumption-a reason that would itself have to transcend their strictly scientific claims and competence 4

Thus while the murkiness of the discussions of naturalism makes it harder to be sure naturalism like closure does not seem to yield an independently defensible presumption in favor of the truth of materialism If there is any better

bull Lurking here is the difficult issue of what sons of entities or properties count as material or physical Is there any good war ro delimit the realm of the material that does not preclude further discoveries in physics but also does not trivialize the categoty by allowing it to include anything that people in departments labeled Physics might eventually come to study This IS anything but a trivial problem but I have no space here to pursue it further

8 9 L Bonjour

rea~n or basis for such a presumption that is prior to and independent of the defense of some particular materialist view I have no idea what it might be

2 FUNCTIONALISM AND CONSCIOUSNESS

The upshot of the previous section is that the case for materialism must rest almost entirely on the defense of particular materialist views and not to any substantial extent on any background presumption So what materialist views are there The answer I think is that once both logical behaviorism and various versions of e1iminativism are set aside as toO implausible to be taken seriously-something that I will assume here without any further discussion-there is only one main materialist view namely functionalism with no very serious prospect that any others will emerge And the fundamental problem for materialism I will suggest is that functionalism offers no account at all of consciousness and seems in principle unable to do so

What gives rise to the mind-body problem in the first place and poses the essential problem that any adequate version of materialism mUSt solve is the fact that conscious mental states as we are aware of them do not present a material appearance-do not seem as we experience them to be material in their makeup in any apparent way Thus a view which holds that everything that exists is material must either (a) deny the very existence of such states as e1iminativism does or else (b) explain how states and correlative properties that do not initially seem to be material in nature can nevertheless turn out to be so A view that takes the latter alternative must give an account of the nature of such states and properties that both accurately reflects their character as experienced and explains how they can nonetheless be entirely material in their makeup And this I suggest is something that has never been successfully done

The starting point for modern versions of materialism was the central-state identity theory particularly the version advocated in a famous paper by J J c Smart (1959) Smart recognized that the truth of materialism can only be an empirical discovery not something knowable a priori For this to be so he argued the various mental states in question must be conceived in a topicshyneutral way a way that makes it possible for them to be merely material in character without implausibly requiring that this be so Only in relation to such a conception would it be possible to discover empirically that such a state is in fact a neurophysiological state of some kind

But for this to work it is crucially important that the topic-neutral conshyception in question be adequate to capture the essential features of mental

Against Materialism

states-something about which Smart was less clear than he might have been For only if this is so will it be the case that showing that the conception offered can be realized by a material state can establish that mental states might in fact be merely material states thereby allowing the rest of the argument to proceed on grounds of simplicity as Smart suggests Whereas if the proposed topic-neutral conception leaves out essential features of mental states-such as consciousness-then the mct that material states can satisfy that conception will be insufficient to explain how mental states might just be material states (Smarts own attempt at a topic-neutral characterization fails to distinguish conscious mental states from whatever else might be going on in the person under a particular set of circumstances)

As in Smarts view functionalism in effect attempts to offer a topic-neutral characterization ofmental states one which will allow but not require that they be essentially material in characters The more general functionalist characterization is in terms of causal role a mental state is characterized by its causal relations to sensory inputs behavioral outputs and other mental states of the same sort The functionalist then proceeds to argue that the states thus characterized could perfectly well be material srates even though the functional characterization does not require rhis A further widely discussed aspect of the view is that different material states could satisfy the functionalist characterization of a particular mental state in different SOrtS ofcreatures or even in the same creature at different times so that (on the most standard version) a material state realizes a functionally characterized state but is not strictly identical with it

But the deepest problem for the functionalist is that the characterization of mental states in terms of causal role say~ nothing at all about consciousness or conscious character There is no apparent reason why a stare thar realizes a particular causal role would thereby need to have any specific sort of conscious character (the point made by the familiar reversed spectrum cases)-or indeed any conscious character at all Thus to point out that a physical state could realize such a causal role really does nothing at all to explain how a conscious state could be (or be realized by) a merely physical state In this way functionalism fails utterly to offer any explanation or account of the most important and conspicuous feature of mental states-or at the very least of a very important and conspicuous feature

It may seem hard to believe that a view that has been held by so many people for so long can be so easily shown to be inadequate in a fundamental way but I think that this is nonetheless so The only solution would be

5 This way of looking at functionalism is explicit in David Lewiss discussion of one of the earliest versions of the view in his 1966 paper An Argument for the Identity Theory Journal of Philosophy 63 17-25 see p 20

10 L Bonjour Against Materialism 11

to offer some supplementary account of what material features give rise to

conscious experience But I know of no such account at least none with any real plausibiliry6

This difficulty with materialism in general and functionalism in particular has of course occasionally been recognized But it still seems to have had remarkably little impact on the widespread acceptance of materialist views I have no very good explanation to offer of this though part of the reason is perhaps the prevailing tendency to approach the philosophy of mind from a third-person neo-behaviorist perspective in which consciousness is largely or entirely ignored (But on this issue it is hard to distinguish the chickens from the eggs)

My basic case against materialism is complete at this point there is no good reason for any strong presumption in favor of materialism and the main materialist view fails to offer any real explanation of a central aspect of mental states namely their conscious character meaning that there is no good reason to think that it is correct as an account of such states But though this very simple argument seems to me entirely compelling I will elaborate it further in the next two sections by focusing on the two main specific kinds of mental states The version of the argument that applies to states with qualitative content is very familiar even though I think that its full force has still not been generally appreciated In contrast the application of essentially the same basic argument to conscious states with intentional content has received far less attention

3 THE PROBLEM OF QUALITATIVE CONSCIOUSNESS

MARY REDUX

Though functionalism fails to adequately account for consciousness of any SOrt

perhaps the most conspicuous aspect ofthis failure penains to qualitative content the sort of content involved in experiences of color and sound and of like pains and itches This point has been made in many ways but the most straightforward and compelling in my view is still the so-called argument initially suggested by Thomas Nagel in relation to the experiences of bats and later developed by Frank Jackson using his famous example of black-and-white Mary on which] will mainly focus here (As most will know

6 One possibiliry is the so-allIed higher order thought theory which holds that consciousne arises when one mental state is the object of a second higher order mental state I have no room here to consider this view in detaiL But the basic-and obvious-problem with such a view is that there is no reason why there could not be a hierarchy of sort even one with many more levels in which there was no consciousness involved at alL (For some elaboration see my contribution to

Bonjour and Sos (2003) pp 65-8) 7 See for example Colin McGinn (1989) David Chalmers (1995) and David Chalmers (1996)

Jackson has changed his mind about this argument and now rejects it though his reasons seem to me unpersuasive)

I will assume here that Jacksons original version of the saga of Mary is familiar enough to require only a brief summation Mary is a brilliant neurophysiologist who lives her entire life acquires her education and does all of her scientific work in a black-and-white environment using black-and-white books and blackshyand-white television for all of her learning and research In this way we may suppose she comes to have a complete knowledge of all the physical facts in neurophysiology and related fields together with their deductive consequences insofar as these are relevanr-thus arriving at as complete an understanding of human functioning as those sciences can provide In particular Mary knows the functional roles of all of the various neurophysiological states including those pertaining to visual perception by knowing their causal relations to sensory inputs behavioral outputs and other such states But despite all of this knowledge Mary apparently does nor know all that there is to know about human mental states for when she is released from her black-and-white environmenr and allowed to view the world normally she will by viewing objects like ripe tomatoes learn what it is like to see something red and analogous things about other qualitative experiences But then comments Jackson it is inescapable that her previous knowledge was incomplete But she had all the physical information Ergo there is more to have than that and Physicalism is falses

Despite the initial force of this rather simple argumenr materialists have nOt been persuaded and the literature comprising materialist responses to the Mary example is very large 9 One thing to say about these responses is that few if any of them are even claimed ro have any substanrial independent plausibility instead they are PUt forward in a way that takes for granted the sort of general presumption in favor of materialism and correlative burden of proof for antishymaterialist views that J have argued does not genuinely exist A full discussion of these responses is impossible here but there are some main themes that can be

dealt with in a general way 10 One of these is the suggestion that although something new when she leaves the black-and-white

room what she acquires is not a knowledge of a new foct (or facts) bur rather something else A second is the suggestion that what she does something like a new ability perhaps more specificall representational ability And if these rwo themes are

bull Frank Jackson (1982) p 130 see also Frank Jackson (1986) bull Many of these discussions are collected in Ludlow Nagasawa and

)0 For a useful taxonomy of the various possible materialist responses see Robert Van Gulick So Many Ways of Saying No to Mary in Ludlow Nagasawa and Stoljar (2004) pp 365-405 (This is of the places where the materialist discussion bears a striking similariry to scholastic theology one can easily imagine a complacent theist writing an artide entitled So Many Ways to Answer the Problem of Evil)

12 L Bonjour i Against Materialism 13

fresult is that there is nothing about the Mary example that is incompatible with Though we have made things vastly easier for Mary by focusing on twO cases the truth of materialism involving colors with which she is now familiar rather than asking her to figure

Particularly in light of rhe general materialist failure to provide an account out on the basis ofher overall physicalneurophysiologicalfunctional knowledge I of conscious experience I doubt very much whether any response of this son would seem even mildly convincing to anyone who was not already determined to adhere to materialism come what may But the first of these fWO themes does at least point to a kind of lacuna in Jacksons original account of the case if Mary learns new facts what exactly are they Indeed in his response to an early version of this suggestion Jackson is reduced to invoking the problem of other minds as a (very) indirect basis for thinking that factual knowledge of some son is involved II

It is however surprisingly easy to modifY the original case in a way that makes it utterly dear that there are facts that Mary does not know while she is in the black-and-white room and will learn when she emerges Suppose that while she is still in the otherwise black-and-white environment fWO color samples are brought in one a sample of a fairly bright green approximately the color of newly mown grass and the other a sample of a fairly bright red approximately the color of a fire engine Mary is allowed to view these samples and even to know that they are fWO of the colors that she has learned about in her black-and-white education She is not however told the standard names of these colors nor is she allowed to monitor her own neurophysiology as she views them

We now remind Mary of two specific ca~es that she has studied morougIllV and about which she knows all the physic~l One of these is a case where a person was second is a case where a person was lookmg at a tell Mary that one of these people had an experience predominar one of the colors with which she is now familiar and that the other person had an experience predominantly involving the other color but of course not which was which If we call the colors presented by the samples color A and color B Mary now knows that one of the two following pairs of claiflls is true

(l) The experience of freshly mown grass predominantly involves color A and the experience of a newly painted fire engine predominantly involve~ color B

(2) The experience of freshly mown grass predominantly involves color B and the experience ofa newly painted fire engine predominantly involves color A

But can she tell on the basis of her black-and-white knowledge together with her new familiarity with the two colors whether it is (1) or (2) which is true (Notice carefully that there is no apparent problem with her understanding of these claims)

l See Jackson (1986 294)

what color experiences in general are like it still seems quite dear for essentially the same reasons that were operative in the original case that she will have no more success with her much more limired task Just as there was nothing in the physical account that could tell her what an experience of red was like so there is still nothing in the physical account of the fire engine case that could definitively pick out the color of one of the samples as opposed to the other 12

And yet whichever of (1) and (2) is true states a foct (or facts) in as robust a sense as one could want-a fact that Mary will learn when she emerges from the black-and-white room and is allowed to view ordinary objects of various kinds

Moreover if there are abilities that result from experiencing the two colors in question Mary presumably can acquire them on the basis of the samples Consider for example Harmans suggestion l3 that what Mary acquires in the original case when she leaves the black-and-white room and sees red for the first time is a perceptual concept of red one that essentially involves being disposed to form perceptual representations involving it in the presence ofcausal stimulation of the right SOrt-sO that she cannot acquire that concept in the original version of the black-and-white room There is much that is questionable about the idea ofsuch a concept but if there is indeed such a thing then Mary in the new version of the case presumably can acquire it by viewing the red sample

more than one sample is for some reason required but it would be easy the new version of the case to allow for that) So we may

suppose Mary has the perceptual concept of red and the perceptual concept of green but she still cannot figure our from her physical knowledge which of these concepts is being employed by the people in the cases she has studied Yet this too is a fact and ifmaterialism isrrue an entirely physical fact So why cant she know it

Here as far as I can see there are only fWO possible moves for the materialist which are even marginally worth considering One is the suggestion that Mary already knows the facts in question as a part of her overall physical knowledge but that she knows them under a different guise or mode of presentation than that under which she will come to know them when she leaves the blackshyand-white room This idea can be developed in different ways and with enormous technical ingenuity But does it really have any serious plausibility Imagine that

J2 As Jackson emphasizes in Jackson (1986 295) it is nO enough for Mary (0 be able to conjecture or guess at the answerO this question For physicalism to be true the fact in question must actually be contained in her physical knowledge

U Gilbert Harman (1990) The Intrinsic Quality of Experience Philosophical Pmpectives 4 3J~52 at pp 44~5 Harman does not actually melllion the Maty case as sucn focusing instead on a person who is blind from birth but still learns all the physical and functional facts of color perception But he does cite Jackson (along with Nagel) as the source of tne objection he is discussing

14 15 L Bonjour

Mary in our modified version of the case having finally experienced real colors is eager to find out more about these intriguing features of the world about which she has been kept in ignorance She wants very much for example to know whether it is (I) or (2) that captures the relevant facts about cases of that sort-and is seriously frustrated about being kept in ignorance any longer Suppose that we respond to her frustration by informing her that she already knows the very facts that she is so eager to learn Surely she would not be satisfied How might she respond

I think we can imagine Mary saying something like this

You philosophers are really amazing The idea that I ar(atiy know the facts I am interested in-indeed all facts of that general kind-is simply preposterous I know all of the physical details but none of them tells me which of the properties I have just experienced on the basis of the samples is realized in each of the two cases If you suggest to me that there arent really novel properties but rather novel concepts or ways of representing or whatever then (while finding that suggestion itself pretry hard to swallow) I would still insist that which concept or way of representing is involved in each case is still something that my physical knowledge doesnt give me any clue about Perhaps as you say there is some clever or complicated way in which the things I want to know are related to the physical things I do know-maybe there is even some metaphysically necessary connection between them (assuming that it is kosher for materialists to believe in such things) Anything like that however just adds to the list of facts that my physical knowledge doesnt reveal to me I am a scientist and not a philosopher so Im not really sure which conception of a fact is the right one (All of the ones you suggest seem pretty weird) But there is undeniably something that I want to know-something that is true about the world-that cant be learned on the basis of all my physical knowledge And that means that the physical story isnt in fact the whole story

Not surprisingly I think that the response I have imagined for Mary is exactly right-that any way of understanding or individuating facts according to which some piece of Marys physical knowledge and either (1) or (2) above turn out to be formulations of the same fact is a conception of fact that is simply too intuitively implausible to be taken seriously

The other possible materialist response is to gram that Mary will learn new facts when she emerges from the black-and-white room in the modified version of the case but to insist that these are nonetheless still physical facts On this view what the case shows is that it is impossible for Mary to acquire complete physical knowledge in the black-and-white room One way to put it is to say that while she can learn all the objective physical facts there are still certain subjective physical facts l4 that she cant learn One can learn what it feels like subjectively to be an organism ofsuch-and-such a general physical description in such-and-such a specific physical state only by actually realizing that condition But that it feels

14 See Van Gulick (2004) for one version of this suggestion

Agaimt Materialism

a certain way or involves a certain sort of conscious experience is still on this view an entirely physical fact

I have to admit that I find it nearly impossible to take this response seriously The only argument for it seems to be an appeal to a background presumption in favor of materialism that is so strong as to make it allegedly reasonable to claim that any fact there is must be a material fact even ifwe cant see in any clear way how it could be a material fact Materialism as we have already seen offers no real account or explanation of consciousness and so also no reason for thinking that there is any subjective experience at all involved in being in a certain material state Thus to advance a view of this sort is in effect just to insist that no fact of any sort can be allowed to refute materialism and thus that any possibiliry of this sort must simply be absorbed into the materialist view however inexplicable in materialist terms it may be (It is not much of a stretch to imagine the materialist saying that we must first believe in order than we may understand)

Thus the modified version of the Mary case seems to present an objection to materialism in general (and functionalism in particular) that is about as conclusive as philosophical arguments ever get However exactly they should be characterized there are facts that Mary cannot know on the basis ofher complete physicalneurophysiologicalfunctional knowledge even when she is given the sort of limited experience needed to understand the claims in question and to acquire any abilities that might be relevant These facts do not seem to be material facts and there is no basis that is not utterly arbitrary and question-begging for supposing that they are Thus we have the strongest of reasons for holding that the materialist account of realiry is incomplete and hence that materialism is

false

4 THE PROBLEM OF CONSCIOUS INTENTIONAL

CONTENT

Qualia of the son involved in the Mary case are widely recognized to pose a serious problem for materialist views and it is not too hard to discern occasional misgivings in this respect under the faltade of materialist confidence But as already mentioned above it is widely assumed that materialism is in much better shape with regard to intentional mental states propositional attitudes and other states that involve aboutness I believe however that this is almost entirely an illusion-that the problems for materialism are just as serious in this area with consciousness being once again the central focus

Materialist accounts of intentional states tend to focus mainly on dispositional states such as beliefs and desires Given the central role of such states in explanations of behavior this is in some ways reasonable enough But such a focus tends to neglect or even ignore the existence of comcious intentional states-even though having conscious thoughts that P is surely one of the

16 L Bonjour

central things that having a dispositional belief that P disposes one to do A focus on belief in particular also has the unfortunate effect of making externalist accounts of intentional content seem more plausible than they possibly could if the emphasis were on conscious intentional states

For these reasons I will focus here explicitly on conscious thoughts As I sit writing this chapter a variety of conscious thoughts pass through my mind Many of these involve the assertion or endorsement of various propositions that materialism cannot account for consciousness that the trees outside my window are very bare that the weather looks cold and dank that the situation in the Middle East looks grim and so on Other thoughts are also propositional but in a way that does not involve assertion my conscious desire to get several pages written before lunch the hope that the stock market will continue to rise and so on It is doubtful that conscious thought must always be propositional in character but it will in any case simplify the issues to be discussed if we largely ignore the propositional aspect of these various thoughts and focus simply on their being conscious thoughts of or about various things or kinds of things materialism the trees the Middle East the stock market and so on

One crucial feature of such conscious thoughts is that when I have them I am in general consciously aware of or consciously understand or grasp what it is that I am thinking about (and also what I am thinking about it) When I think that the trees outside my window are bare I consciously understand that it is certain trees that I am thinking about (and along with this what sort of thing a tree is and which trees I have in mind) What exactly this conscious grasp of the object of thought involves varies from case to case and is sometimes not easy to precisely specify Moreover as will emerge it is something of which I think we presently have no real explanatory account of any substance But its existence is I submit completely undeniable Indeed being able in this way to consciously think about things to have them in mind is in many ways the most central and obvious feature of our mental lives

It is obvious that a persons conscious grasp of the object of their thought of what they are thinking about can vary on a number of dimensions it may be more or less precise more or less detailed more or less clear more or less complete But contrary to what is sometimes suggested it is rarely if ever merely disquotational in character Perhaps (though I doubt it) there are cases where a scientifically untutored person is thinking about eg electrons and where their sole grasp of what they are thinking is that it is what is referred to in their society or community by the word electron -so that what they are thinking about is in effect electrons (whatever they are) But this is surely not the ordinary situation when we think about various things 15

15 Notice that even a thought about what the relevant Ioci~tal ocp~rts mean by electrons would have to involve a non-disquotational element in the reference to those experts and also in the reference to the word to think about whatever it is that the societal experts mean by electrons

Against Materialism 17

Moreover the existence of conscious intentional content is perfectly compatshyible with the existence of an externalist dimension of thought content-though not with the view that all content is external If as in Putnams famous example a person is thinking about earthly water at a time prior to the discovery of its chemical composition there is no reason to deny that they are in a sense thinking about H 20 But in such a situation the aspect of being about H 20 will obviously not be part of their conscious internal grasp of what they are thinking about in the way that the more superficial aspects of water will 16 And in a somewhat parallel way the person in Burges famous example who thinks he has arthritis in his thigh and to whom our standard belief-ascription practices ascribe a belief about arthritis (where this is among other things a disease that only occurs in joints) obviously does not consciously grasp the disease that he is thinking about in a way that involves this specific feature of itY But it is nonetheless impossible to describe either example in a convincing way without presupposing that the people in question do have something consciously in mind a substance having the superficial properties of water in Putnams example and a disease having certain fairly specific features in Burges Thus while it is possible to dispute the relative importance ofconscious internal thought content and external thought content of which the subject is not conscious examples of this sort provide no basis at all for denying that conscious internal content exists

The issue I want to raise here is whether a materialist view can account for the sort of conscious intentional content just characterized Can it account for conscious thoughts being about various things in a way that can be grasped or understood by the person in question In a way the answer has already been given Since materialist views really take no account at all of consciousness they obviously offer no account of this particular aspect of it But investigating this narrower aspect of the issue can still help to deepen the basic objection to

materialism Here it will be useful to bring the brilliant neurophysiologist Mary briefly

back onto the scene even though the black-and-white aspect of her situation is no longer relevant Suppose that Mary studies me as a subject and comes to

have a complete knowledge of my physical and neurophysiological makeup as I am thinking these various thoughts Can she determine on that basis what I am consciously thinking about at a particular moment

One thing that seems utterly clear is that she could not do this merely on the basis of knowing my internal physical characteristics-as it is sometimes put

I knowing everything physical that happens inside my skin There is no reason t f

is not the same thing as thinking about whatever is meant by electrons by whatever is meant by the societal experts

I 16 See Hilary Putnam (1975a) (Cambridge Cambridge Universiry Press) pp 215-71 17 See Tyler Burge (1979)

19 18 L Bonjour

at all to think that the internal structure of my physical and neurophysiological states could somehow by itself determine that I am thinking about weather tather than about the Middle East or the stock market

A functionalist would no doubt say that it is no surprise that Mary could not do this In order to know the complete causal or functional role of my internal states Mary also needs to know about their external causal relations to various things And it might be suggested if Mary knows all of the external causal relations in which my various states stand she will in fact be able ro figure out what I am consciously thinking about at any particular time No doubt the details that pick out any particular object of thought will be very complicated but there is it might be claimed no reason to doubt that in principle she could do this

Here we have a piece of materialist doctrine that again has a status very similar to that of a claim of theology It is obvious that no one ha~ even the beginnings of an idea of how to actually carry out an investigation that would yield a result of this kind-that the only rea~on for thinking that this could be done is the overriding assumption for which we have found no cogent basis that materialism must be true Among a multitude of other difficulties Mary would have to be able to figure out the content of thoughts that are confused or inaccurate or thoughts about imaginary or fictional entities or supernatural entities It is to say the least very hard to see how she could do this on the basis of a knowledge of causal relations to more ordinary sorts of things

But the problem for materialism is in fact even worse than that For as already emphasized it is an undeniable fact about conscious intentional content that I am able for the most part to consciously understand or be aware of what I am thinking about from the inside Clearly I do not in general do this on the basis of external causal knowledge I do not have such knowledge and would not know what to do about it if I did All that I normally have any sort of direct access to if materialism is true is my own internal physical and physiological states and thus my conscious understanding of what I am thinking about at a particular moment must be somehow a feature or result of those internal states alone Causal relations to external things may help to produce the relevant features of the internal states in question but there is no apparent way in which such external relations can somehow be partly constitutive of the fact that my conscious thoughts are about various things in a way of which I can be immediately aware But if these internal states are sufficient to fix the object of my thought in a way that is accessible to my understanding or awareness then knowing about those internal states should be sufficient for Mary a~ well without any knowledge of the external causal relations And yet as we have already seen it seems obvious that this is not the ca~e18

18 Ir is worth noting that the same thing is really true in the ase of qualia as well A persons awareness of one color rather than another when he or she looks at newly mown grass obviously

Against Materialism

Thus we have the basis for an argument that is parallel to Jacksons original argument about qualia Mary knows all the relevant physical facts she is not able on the basis of this knowledge to know what I am consciously thinking about at a particular moment but what I am thinking about at that moment is as surely a fact about the world as anything else therefore complete physical knowledge is not complete knowledge and so materialism is false

One way to further elaborate this point is to consider how it applies to

what is perhaps the most widely held materialist view of intentional content the view popularized by Jerry Fodor and many others that intentional mental states employ an internal language a language of thought19 Fodor calls this view the representational theory of the mind though it might better be called the symbolic theory of the mind For the crucial feature of the view is that the language of thought like any language is composed of symbols items that do not stand for anything by virtue of their intrinsic properties but whose representative character depends instead on the relations in which they stand to other things-for Fodor the sorts of causal relations that are captured in the idea of a causal or functional role1o Just as the word dog could in principle have stood for anything (or nothing at all) and in fact stands for a kind ofanimal rather than something else only because of causal relations that arise from the way it is used so also the symbols in the language of thought stand for whatever they stand for only by virrue of analogous sorts of relations and not in virtue of their intrinsic physical and neurophysiological properties Their intentional character is thus extrinsic not intrinsic

Proponents of the language ofthought rarely have much to say about conscious thoughts of the sort that we are focusing on here But it is clear that on their view what happens when I am consciously thinking about say the Middle East is that in some appropriate location in my overall cognitive operations there occurs a symbol (or set of symbols) that refers to the Middle East This symbol

f like the surrounding context in which it occurs is some neurophysiological state or some constellation of such states No one of course has at present any real

knowledge of the concrete nature of such symbols or their larger contexts but it f I will do no harm to follow Fodor in thinking of a mental blackboard on which

mental symbols are inscribed in appropriate ways Thus for me ro be consciously thinking about the Middle East is for me ro have the mental symbol that refers to

does not depend in a constitutive way on external causal rdations even though it may be causally produced by them Thus in that case too a knowledge of the persons internal physical and neurophysiological states alone should enable Mary to pick out One color rather than the other as the right one But it is even more obvious than in the original ase that this is not so

See eg Jerry Fodor Propositional Attitudes and Methodological Solipsism Considered as a Research Srrategy in Cognitive Science both reprinted in his 198 book Rtpwmll1rions (Cambridge MA MIT Press)

20 See for example Jerry Fodor chapter 4 Fodor has subsequently refined this view in various ways but none that affect issues being raised here

20 L Bonjour Against Materialism 21

the Middle East inscribed in the right way on this blackboard But the symbols reference to the Middle East to repeat depends not on its intrinsic physical or neurophysiological character alone but also on the relations in which it stands to other such symbols and directly or indirectly to the external world

Suppose now that Mary is studying my cognitive operations Suppose that she has somehow isolated what amounts to my mental blackboard and the various symbols written on it Obviously this will not in itself tell her what I am thinking about Even if she could somehow focus onthe specific symbol that refers to the Middle East and tell that it is functioning in a way that determines the object of my conscious thought (even though there is no reason to think that she could in fact do these things) she will not on this basis alone be able to tell what it is that this symbol in fact refers to Nor is there any plausibiliry to the idea that Mary could figure out the reference or meaning of the various mental symbols simply by examining their internal relations (0 each other21 Thus she will need once again to appeal to external causal relations ofvarious sorts

Bur how then am I able to be aware ofor understand from the inside what I am thinking about Once again I have no knowledge of those external relations (and would be very unlikely to be able to figure anything out from them even if I did) All that I plausibly have access to is the mental symbol or symbols and the surrounding system of states and this is apparently nor enough to determine the object of my thought

The only very obvious recourse here for the proponent of a language of thought is to construe my understanding or awareness of what I am thinking about disquotationally in relation to the language of thought When thinking about the Middle East I do so by using some mental symbol And when [ understand or am aware of what I am thinking about it might be suggested I in effect use that very same symbol what I am aware of is that I am thinking about the Middle East (whatever that is-that is whatever that symbol in fact refers (0) If the symbol in question did succeed in referring to the Middle East then this specification of what I am thinking about will refer to the Middle East as well and so will be correct But it is intuitively as obvious as anything could be that my awareness ofwhat I am thinking about normally involves more than this involves actually understanding (at some level of precision detail etc) what the Middle East is in a way that goes beyond merely repeating the same symbol Assuming for the moment that there really is a language of thought I understand my language of thought in a way parallel to the way in which I understand my own public language-and not in the merely disquotational way that could just as well be applied to a language of which I have no understanding at all

Here a proponent of the language of thought may want to reply that the difference in the public language case is merely that one language is a language

2 Such an idea has sometimes at least apparently been suggested For more discussion see 01)

19()8 book In Difmse ofPure Reason pp 174-80

that I successfully use-and that the same is true of my language of thought On this view the intuition that I understand what I am thinking about-or what I am talking about-in any stronger sense one that is not merely disquotational is merely an illusion But here again we have a view that it seems to me would appeal to no one who was not motivated by the conviction that materialism must be true

My conclusion is that the language of thought view has nothing useful to say about the most obvious sort of intentional content the intentional content that is involved in having something explicitly and consciously in mind Nor do I know of any other materialist account that does any better in this regard There is perhaps room for dispute about just how important conscious intentional content is in relation to the causation and explanation of behavior but no plausible way (0 deny that it genuinely exists Thus with respect to intentional content as with the case of qualitative content materialism seems to be utterly bankrupt as a general account of mental states and to be held merely as an article of faith

5 WHAT IS THE ALTERNATIVE

The last two sections serve merely to strengthen and deepen the fundamental objection to materialism already offered in section 2 consciousness genuinely exists materialism can oller no account that explains consciousness (or of the specific varieties thereof) or shows it [0 be merely material in character therefore (at least in the absence of any strong antecedent argument or presumption in favor of materialism) the indicated conclusion is that materialism is false There is more in heaven and earth than is dreamt of in materialist philosophy

But what do I mean by more Here as I see it there is very little that can be said in our current state of knowledge so thar the main result is that we have very little understanding ofconsciousness-or given the arguably central role of consciousness of mentaliry in general

In the first place there is no clear way in which the objections that I have raised against materialism support the classical substance dualist position Positing a separate mental substance that is characterized in almost entirely negative terms does nothing very obvious to explain consciousness in general or qualitative and intentional content in particular As far as I can see the main appeal of substance dualism is that the account of the supposed mental or spiritual substances is fur too vague and sketchy to provide the basis for any very clear argument that such substances could not be the locus of consciousness But this negative point hardly counts as an argument in favor of such a view

The obvious alternative is properry dualism the view that human persons and perhaps other kinds of animals have non-material or non-physical properties in addition to their physical ones with at least the main such properties being

8 9 L Bonjour

rea~n or basis for such a presumption that is prior to and independent of the defense of some particular materialist view I have no idea what it might be

2 FUNCTIONALISM AND CONSCIOUSNESS

The upshot of the previous section is that the case for materialism must rest almost entirely on the defense of particular materialist views and not to any substantial extent on any background presumption So what materialist views are there The answer I think is that once both logical behaviorism and various versions of e1iminativism are set aside as toO implausible to be taken seriously-something that I will assume here without any further discussion-there is only one main materialist view namely functionalism with no very serious prospect that any others will emerge And the fundamental problem for materialism I will suggest is that functionalism offers no account at all of consciousness and seems in principle unable to do so

What gives rise to the mind-body problem in the first place and poses the essential problem that any adequate version of materialism mUSt solve is the fact that conscious mental states as we are aware of them do not present a material appearance-do not seem as we experience them to be material in their makeup in any apparent way Thus a view which holds that everything that exists is material must either (a) deny the very existence of such states as e1iminativism does or else (b) explain how states and correlative properties that do not initially seem to be material in nature can nevertheless turn out to be so A view that takes the latter alternative must give an account of the nature of such states and properties that both accurately reflects their character as experienced and explains how they can nonetheless be entirely material in their makeup And this I suggest is something that has never been successfully done

The starting point for modern versions of materialism was the central-state identity theory particularly the version advocated in a famous paper by J J c Smart (1959) Smart recognized that the truth of materialism can only be an empirical discovery not something knowable a priori For this to be so he argued the various mental states in question must be conceived in a topicshyneutral way a way that makes it possible for them to be merely material in character without implausibly requiring that this be so Only in relation to such a conception would it be possible to discover empirically that such a state is in fact a neurophysiological state of some kind

But for this to work it is crucially important that the topic-neutral conshyception in question be adequate to capture the essential features of mental

Against Materialism

states-something about which Smart was less clear than he might have been For only if this is so will it be the case that showing that the conception offered can be realized by a material state can establish that mental states might in fact be merely material states thereby allowing the rest of the argument to proceed on grounds of simplicity as Smart suggests Whereas if the proposed topic-neutral conception leaves out essential features of mental states-such as consciousness-then the mct that material states can satisfy that conception will be insufficient to explain how mental states might just be material states (Smarts own attempt at a topic-neutral characterization fails to distinguish conscious mental states from whatever else might be going on in the person under a particular set of circumstances)

As in Smarts view functionalism in effect attempts to offer a topic-neutral characterization ofmental states one which will allow but not require that they be essentially material in characters The more general functionalist characterization is in terms of causal role a mental state is characterized by its causal relations to sensory inputs behavioral outputs and other mental states of the same sort The functionalist then proceeds to argue that the states thus characterized could perfectly well be material srates even though the functional characterization does not require rhis A further widely discussed aspect of the view is that different material states could satisfy the functionalist characterization of a particular mental state in different SOrtS ofcreatures or even in the same creature at different times so that (on the most standard version) a material state realizes a functionally characterized state but is not strictly identical with it

But the deepest problem for the functionalist is that the characterization of mental states in terms of causal role say~ nothing at all about consciousness or conscious character There is no apparent reason why a stare thar realizes a particular causal role would thereby need to have any specific sort of conscious character (the point made by the familiar reversed spectrum cases)-or indeed any conscious character at all Thus to point out that a physical state could realize such a causal role really does nothing at all to explain how a conscious state could be (or be realized by) a merely physical state In this way functionalism fails utterly to offer any explanation or account of the most important and conspicuous feature of mental states-or at the very least of a very important and conspicuous feature

It may seem hard to believe that a view that has been held by so many people for so long can be so easily shown to be inadequate in a fundamental way but I think that this is nonetheless so The only solution would be

5 This way of looking at functionalism is explicit in David Lewiss discussion of one of the earliest versions of the view in his 1966 paper An Argument for the Identity Theory Journal of Philosophy 63 17-25 see p 20

10 L Bonjour Against Materialism 11

to offer some supplementary account of what material features give rise to

conscious experience But I know of no such account at least none with any real plausibiliry6

This difficulty with materialism in general and functionalism in particular has of course occasionally been recognized But it still seems to have had remarkably little impact on the widespread acceptance of materialist views I have no very good explanation to offer of this though part of the reason is perhaps the prevailing tendency to approach the philosophy of mind from a third-person neo-behaviorist perspective in which consciousness is largely or entirely ignored (But on this issue it is hard to distinguish the chickens from the eggs)

My basic case against materialism is complete at this point there is no good reason for any strong presumption in favor of materialism and the main materialist view fails to offer any real explanation of a central aspect of mental states namely their conscious character meaning that there is no good reason to think that it is correct as an account of such states But though this very simple argument seems to me entirely compelling I will elaborate it further in the next two sections by focusing on the two main specific kinds of mental states The version of the argument that applies to states with qualitative content is very familiar even though I think that its full force has still not been generally appreciated In contrast the application of essentially the same basic argument to conscious states with intentional content has received far less attention

3 THE PROBLEM OF QUALITATIVE CONSCIOUSNESS

MARY REDUX

Though functionalism fails to adequately account for consciousness of any SOrt

perhaps the most conspicuous aspect ofthis failure penains to qualitative content the sort of content involved in experiences of color and sound and of like pains and itches This point has been made in many ways but the most straightforward and compelling in my view is still the so-called argument initially suggested by Thomas Nagel in relation to the experiences of bats and later developed by Frank Jackson using his famous example of black-and-white Mary on which] will mainly focus here (As most will know

6 One possibiliry is the so-allIed higher order thought theory which holds that consciousne arises when one mental state is the object of a second higher order mental state I have no room here to consider this view in detaiL But the basic-and obvious-problem with such a view is that there is no reason why there could not be a hierarchy of sort even one with many more levels in which there was no consciousness involved at alL (For some elaboration see my contribution to

Bonjour and Sos (2003) pp 65-8) 7 See for example Colin McGinn (1989) David Chalmers (1995) and David Chalmers (1996)

Jackson has changed his mind about this argument and now rejects it though his reasons seem to me unpersuasive)

I will assume here that Jacksons original version of the saga of Mary is familiar enough to require only a brief summation Mary is a brilliant neurophysiologist who lives her entire life acquires her education and does all of her scientific work in a black-and-white environment using black-and-white books and blackshyand-white television for all of her learning and research In this way we may suppose she comes to have a complete knowledge of all the physical facts in neurophysiology and related fields together with their deductive consequences insofar as these are relevanr-thus arriving at as complete an understanding of human functioning as those sciences can provide In particular Mary knows the functional roles of all of the various neurophysiological states including those pertaining to visual perception by knowing their causal relations to sensory inputs behavioral outputs and other such states But despite all of this knowledge Mary apparently does nor know all that there is to know about human mental states for when she is released from her black-and-white environmenr and allowed to view the world normally she will by viewing objects like ripe tomatoes learn what it is like to see something red and analogous things about other qualitative experiences But then comments Jackson it is inescapable that her previous knowledge was incomplete But she had all the physical information Ergo there is more to have than that and Physicalism is falses

Despite the initial force of this rather simple argumenr materialists have nOt been persuaded and the literature comprising materialist responses to the Mary example is very large 9 One thing to say about these responses is that few if any of them are even claimed ro have any substanrial independent plausibility instead they are PUt forward in a way that takes for granted the sort of general presumption in favor of materialism and correlative burden of proof for antishymaterialist views that J have argued does not genuinely exist A full discussion of these responses is impossible here but there are some main themes that can be

dealt with in a general way 10 One of these is the suggestion that although something new when she leaves the black-and-white

room what she acquires is not a knowledge of a new foct (or facts) bur rather something else A second is the suggestion that what she does something like a new ability perhaps more specificall representational ability And if these rwo themes are

bull Frank Jackson (1982) p 130 see also Frank Jackson (1986) bull Many of these discussions are collected in Ludlow Nagasawa and

)0 For a useful taxonomy of the various possible materialist responses see Robert Van Gulick So Many Ways of Saying No to Mary in Ludlow Nagasawa and Stoljar (2004) pp 365-405 (This is of the places where the materialist discussion bears a striking similariry to scholastic theology one can easily imagine a complacent theist writing an artide entitled So Many Ways to Answer the Problem of Evil)

12 L Bonjour i Against Materialism 13

fresult is that there is nothing about the Mary example that is incompatible with Though we have made things vastly easier for Mary by focusing on twO cases the truth of materialism involving colors with which she is now familiar rather than asking her to figure

Particularly in light of rhe general materialist failure to provide an account out on the basis ofher overall physicalneurophysiologicalfunctional knowledge I of conscious experience I doubt very much whether any response of this son would seem even mildly convincing to anyone who was not already determined to adhere to materialism come what may But the first of these fWO themes does at least point to a kind of lacuna in Jacksons original account of the case if Mary learns new facts what exactly are they Indeed in his response to an early version of this suggestion Jackson is reduced to invoking the problem of other minds as a (very) indirect basis for thinking that factual knowledge of some son is involved II

It is however surprisingly easy to modifY the original case in a way that makes it utterly dear that there are facts that Mary does not know while she is in the black-and-white room and will learn when she emerges Suppose that while she is still in the otherwise black-and-white environment fWO color samples are brought in one a sample of a fairly bright green approximately the color of newly mown grass and the other a sample of a fairly bright red approximately the color of a fire engine Mary is allowed to view these samples and even to know that they are fWO of the colors that she has learned about in her black-and-white education She is not however told the standard names of these colors nor is she allowed to monitor her own neurophysiology as she views them

We now remind Mary of two specific ca~es that she has studied morougIllV and about which she knows all the physic~l One of these is a case where a person was second is a case where a person was lookmg at a tell Mary that one of these people had an experience predominar one of the colors with which she is now familiar and that the other person had an experience predominantly involving the other color but of course not which was which If we call the colors presented by the samples color A and color B Mary now knows that one of the two following pairs of claiflls is true

(l) The experience of freshly mown grass predominantly involves color A and the experience of a newly painted fire engine predominantly involve~ color B

(2) The experience of freshly mown grass predominantly involves color B and the experience ofa newly painted fire engine predominantly involves color A

But can she tell on the basis of her black-and-white knowledge together with her new familiarity with the two colors whether it is (1) or (2) which is true (Notice carefully that there is no apparent problem with her understanding of these claims)

l See Jackson (1986 294)

what color experiences in general are like it still seems quite dear for essentially the same reasons that were operative in the original case that she will have no more success with her much more limired task Just as there was nothing in the physical account that could tell her what an experience of red was like so there is still nothing in the physical account of the fire engine case that could definitively pick out the color of one of the samples as opposed to the other 12

And yet whichever of (1) and (2) is true states a foct (or facts) in as robust a sense as one could want-a fact that Mary will learn when she emerges from the black-and-white room and is allowed to view ordinary objects of various kinds

Moreover if there are abilities that result from experiencing the two colors in question Mary presumably can acquire them on the basis of the samples Consider for example Harmans suggestion l3 that what Mary acquires in the original case when she leaves the black-and-white room and sees red for the first time is a perceptual concept of red one that essentially involves being disposed to form perceptual representations involving it in the presence ofcausal stimulation of the right SOrt-sO that she cannot acquire that concept in the original version of the black-and-white room There is much that is questionable about the idea ofsuch a concept but if there is indeed such a thing then Mary in the new version of the case presumably can acquire it by viewing the red sample

more than one sample is for some reason required but it would be easy the new version of the case to allow for that) So we may

suppose Mary has the perceptual concept of red and the perceptual concept of green but she still cannot figure our from her physical knowledge which of these concepts is being employed by the people in the cases she has studied Yet this too is a fact and ifmaterialism isrrue an entirely physical fact So why cant she know it

Here as far as I can see there are only fWO possible moves for the materialist which are even marginally worth considering One is the suggestion that Mary already knows the facts in question as a part of her overall physical knowledge but that she knows them under a different guise or mode of presentation than that under which she will come to know them when she leaves the blackshyand-white room This idea can be developed in different ways and with enormous technical ingenuity But does it really have any serious plausibility Imagine that

J2 As Jackson emphasizes in Jackson (1986 295) it is nO enough for Mary (0 be able to conjecture or guess at the answerO this question For physicalism to be true the fact in question must actually be contained in her physical knowledge

U Gilbert Harman (1990) The Intrinsic Quality of Experience Philosophical Pmpectives 4 3J~52 at pp 44~5 Harman does not actually melllion the Maty case as sucn focusing instead on a person who is blind from birth but still learns all the physical and functional facts of color perception But he does cite Jackson (along with Nagel) as the source of tne objection he is discussing

14 15 L Bonjour

Mary in our modified version of the case having finally experienced real colors is eager to find out more about these intriguing features of the world about which she has been kept in ignorance She wants very much for example to know whether it is (I) or (2) that captures the relevant facts about cases of that sort-and is seriously frustrated about being kept in ignorance any longer Suppose that we respond to her frustration by informing her that she already knows the very facts that she is so eager to learn Surely she would not be satisfied How might she respond

I think we can imagine Mary saying something like this

You philosophers are really amazing The idea that I ar(atiy know the facts I am interested in-indeed all facts of that general kind-is simply preposterous I know all of the physical details but none of them tells me which of the properties I have just experienced on the basis of the samples is realized in each of the two cases If you suggest to me that there arent really novel properties but rather novel concepts or ways of representing or whatever then (while finding that suggestion itself pretry hard to swallow) I would still insist that which concept or way of representing is involved in each case is still something that my physical knowledge doesnt give me any clue about Perhaps as you say there is some clever or complicated way in which the things I want to know are related to the physical things I do know-maybe there is even some metaphysically necessary connection between them (assuming that it is kosher for materialists to believe in such things) Anything like that however just adds to the list of facts that my physical knowledge doesnt reveal to me I am a scientist and not a philosopher so Im not really sure which conception of a fact is the right one (All of the ones you suggest seem pretty weird) But there is undeniably something that I want to know-something that is true about the world-that cant be learned on the basis of all my physical knowledge And that means that the physical story isnt in fact the whole story

Not surprisingly I think that the response I have imagined for Mary is exactly right-that any way of understanding or individuating facts according to which some piece of Marys physical knowledge and either (1) or (2) above turn out to be formulations of the same fact is a conception of fact that is simply too intuitively implausible to be taken seriously

The other possible materialist response is to gram that Mary will learn new facts when she emerges from the black-and-white room in the modified version of the case but to insist that these are nonetheless still physical facts On this view what the case shows is that it is impossible for Mary to acquire complete physical knowledge in the black-and-white room One way to put it is to say that while she can learn all the objective physical facts there are still certain subjective physical facts l4 that she cant learn One can learn what it feels like subjectively to be an organism ofsuch-and-such a general physical description in such-and-such a specific physical state only by actually realizing that condition But that it feels

14 See Van Gulick (2004) for one version of this suggestion

Agaimt Materialism

a certain way or involves a certain sort of conscious experience is still on this view an entirely physical fact

I have to admit that I find it nearly impossible to take this response seriously The only argument for it seems to be an appeal to a background presumption in favor of materialism that is so strong as to make it allegedly reasonable to claim that any fact there is must be a material fact even ifwe cant see in any clear way how it could be a material fact Materialism as we have already seen offers no real account or explanation of consciousness and so also no reason for thinking that there is any subjective experience at all involved in being in a certain material state Thus to advance a view of this sort is in effect just to insist that no fact of any sort can be allowed to refute materialism and thus that any possibiliry of this sort must simply be absorbed into the materialist view however inexplicable in materialist terms it may be (It is not much of a stretch to imagine the materialist saying that we must first believe in order than we may understand)

Thus the modified version of the Mary case seems to present an objection to materialism in general (and functionalism in particular) that is about as conclusive as philosophical arguments ever get However exactly they should be characterized there are facts that Mary cannot know on the basis ofher complete physicalneurophysiologicalfunctional knowledge even when she is given the sort of limited experience needed to understand the claims in question and to acquire any abilities that might be relevant These facts do not seem to be material facts and there is no basis that is not utterly arbitrary and question-begging for supposing that they are Thus we have the strongest of reasons for holding that the materialist account of realiry is incomplete and hence that materialism is

false

4 THE PROBLEM OF CONSCIOUS INTENTIONAL

CONTENT

Qualia of the son involved in the Mary case are widely recognized to pose a serious problem for materialist views and it is not too hard to discern occasional misgivings in this respect under the faltade of materialist confidence But as already mentioned above it is widely assumed that materialism is in much better shape with regard to intentional mental states propositional attitudes and other states that involve aboutness I believe however that this is almost entirely an illusion-that the problems for materialism are just as serious in this area with consciousness being once again the central focus

Materialist accounts of intentional states tend to focus mainly on dispositional states such as beliefs and desires Given the central role of such states in explanations of behavior this is in some ways reasonable enough But such a focus tends to neglect or even ignore the existence of comcious intentional states-even though having conscious thoughts that P is surely one of the

16 L Bonjour

central things that having a dispositional belief that P disposes one to do A focus on belief in particular also has the unfortunate effect of making externalist accounts of intentional content seem more plausible than they possibly could if the emphasis were on conscious intentional states

For these reasons I will focus here explicitly on conscious thoughts As I sit writing this chapter a variety of conscious thoughts pass through my mind Many of these involve the assertion or endorsement of various propositions that materialism cannot account for consciousness that the trees outside my window are very bare that the weather looks cold and dank that the situation in the Middle East looks grim and so on Other thoughts are also propositional but in a way that does not involve assertion my conscious desire to get several pages written before lunch the hope that the stock market will continue to rise and so on It is doubtful that conscious thought must always be propositional in character but it will in any case simplify the issues to be discussed if we largely ignore the propositional aspect of these various thoughts and focus simply on their being conscious thoughts of or about various things or kinds of things materialism the trees the Middle East the stock market and so on

One crucial feature of such conscious thoughts is that when I have them I am in general consciously aware of or consciously understand or grasp what it is that I am thinking about (and also what I am thinking about it) When I think that the trees outside my window are bare I consciously understand that it is certain trees that I am thinking about (and along with this what sort of thing a tree is and which trees I have in mind) What exactly this conscious grasp of the object of thought involves varies from case to case and is sometimes not easy to precisely specify Moreover as will emerge it is something of which I think we presently have no real explanatory account of any substance But its existence is I submit completely undeniable Indeed being able in this way to consciously think about things to have them in mind is in many ways the most central and obvious feature of our mental lives

It is obvious that a persons conscious grasp of the object of their thought of what they are thinking about can vary on a number of dimensions it may be more or less precise more or less detailed more or less clear more or less complete But contrary to what is sometimes suggested it is rarely if ever merely disquotational in character Perhaps (though I doubt it) there are cases where a scientifically untutored person is thinking about eg electrons and where their sole grasp of what they are thinking is that it is what is referred to in their society or community by the word electron -so that what they are thinking about is in effect electrons (whatever they are) But this is surely not the ordinary situation when we think about various things 15

15 Notice that even a thought about what the relevant Ioci~tal ocp~rts mean by electrons would have to involve a non-disquotational element in the reference to those experts and also in the reference to the word to think about whatever it is that the societal experts mean by electrons

Against Materialism 17

Moreover the existence of conscious intentional content is perfectly compatshyible with the existence of an externalist dimension of thought content-though not with the view that all content is external If as in Putnams famous example a person is thinking about earthly water at a time prior to the discovery of its chemical composition there is no reason to deny that they are in a sense thinking about H 20 But in such a situation the aspect of being about H 20 will obviously not be part of their conscious internal grasp of what they are thinking about in the way that the more superficial aspects of water will 16 And in a somewhat parallel way the person in Burges famous example who thinks he has arthritis in his thigh and to whom our standard belief-ascription practices ascribe a belief about arthritis (where this is among other things a disease that only occurs in joints) obviously does not consciously grasp the disease that he is thinking about in a way that involves this specific feature of itY But it is nonetheless impossible to describe either example in a convincing way without presupposing that the people in question do have something consciously in mind a substance having the superficial properties of water in Putnams example and a disease having certain fairly specific features in Burges Thus while it is possible to dispute the relative importance ofconscious internal thought content and external thought content of which the subject is not conscious examples of this sort provide no basis at all for denying that conscious internal content exists

The issue I want to raise here is whether a materialist view can account for the sort of conscious intentional content just characterized Can it account for conscious thoughts being about various things in a way that can be grasped or understood by the person in question In a way the answer has already been given Since materialist views really take no account at all of consciousness they obviously offer no account of this particular aspect of it But investigating this narrower aspect of the issue can still help to deepen the basic objection to

materialism Here it will be useful to bring the brilliant neurophysiologist Mary briefly

back onto the scene even though the black-and-white aspect of her situation is no longer relevant Suppose that Mary studies me as a subject and comes to

have a complete knowledge of my physical and neurophysiological makeup as I am thinking these various thoughts Can she determine on that basis what I am consciously thinking about at a particular moment

One thing that seems utterly clear is that she could not do this merely on the basis of knowing my internal physical characteristics-as it is sometimes put

I knowing everything physical that happens inside my skin There is no reason t f

is not the same thing as thinking about whatever is meant by electrons by whatever is meant by the societal experts

I 16 See Hilary Putnam (1975a) (Cambridge Cambridge Universiry Press) pp 215-71 17 See Tyler Burge (1979)

19 18 L Bonjour

at all to think that the internal structure of my physical and neurophysiological states could somehow by itself determine that I am thinking about weather tather than about the Middle East or the stock market

A functionalist would no doubt say that it is no surprise that Mary could not do this In order to know the complete causal or functional role of my internal states Mary also needs to know about their external causal relations to various things And it might be suggested if Mary knows all of the external causal relations in which my various states stand she will in fact be able ro figure out what I am consciously thinking about at any particular time No doubt the details that pick out any particular object of thought will be very complicated but there is it might be claimed no reason to doubt that in principle she could do this

Here we have a piece of materialist doctrine that again has a status very similar to that of a claim of theology It is obvious that no one ha~ even the beginnings of an idea of how to actually carry out an investigation that would yield a result of this kind-that the only rea~on for thinking that this could be done is the overriding assumption for which we have found no cogent basis that materialism must be true Among a multitude of other difficulties Mary would have to be able to figure out the content of thoughts that are confused or inaccurate or thoughts about imaginary or fictional entities or supernatural entities It is to say the least very hard to see how she could do this on the basis of a knowledge of causal relations to more ordinary sorts of things

But the problem for materialism is in fact even worse than that For as already emphasized it is an undeniable fact about conscious intentional content that I am able for the most part to consciously understand or be aware of what I am thinking about from the inside Clearly I do not in general do this on the basis of external causal knowledge I do not have such knowledge and would not know what to do about it if I did All that I normally have any sort of direct access to if materialism is true is my own internal physical and physiological states and thus my conscious understanding of what I am thinking about at a particular moment must be somehow a feature or result of those internal states alone Causal relations to external things may help to produce the relevant features of the internal states in question but there is no apparent way in which such external relations can somehow be partly constitutive of the fact that my conscious thoughts are about various things in a way of which I can be immediately aware But if these internal states are sufficient to fix the object of my thought in a way that is accessible to my understanding or awareness then knowing about those internal states should be sufficient for Mary a~ well without any knowledge of the external causal relations And yet as we have already seen it seems obvious that this is not the ca~e18

18 Ir is worth noting that the same thing is really true in the ase of qualia as well A persons awareness of one color rather than another when he or she looks at newly mown grass obviously

Against Materialism

Thus we have the basis for an argument that is parallel to Jacksons original argument about qualia Mary knows all the relevant physical facts she is not able on the basis of this knowledge to know what I am consciously thinking about at a particular moment but what I am thinking about at that moment is as surely a fact about the world as anything else therefore complete physical knowledge is not complete knowledge and so materialism is false

One way to further elaborate this point is to consider how it applies to

what is perhaps the most widely held materialist view of intentional content the view popularized by Jerry Fodor and many others that intentional mental states employ an internal language a language of thought19 Fodor calls this view the representational theory of the mind though it might better be called the symbolic theory of the mind For the crucial feature of the view is that the language of thought like any language is composed of symbols items that do not stand for anything by virtue of their intrinsic properties but whose representative character depends instead on the relations in which they stand to other things-for Fodor the sorts of causal relations that are captured in the idea of a causal or functional role1o Just as the word dog could in principle have stood for anything (or nothing at all) and in fact stands for a kind ofanimal rather than something else only because of causal relations that arise from the way it is used so also the symbols in the language of thought stand for whatever they stand for only by virrue of analogous sorts of relations and not in virtue of their intrinsic physical and neurophysiological properties Their intentional character is thus extrinsic not intrinsic

Proponents of the language ofthought rarely have much to say about conscious thoughts of the sort that we are focusing on here But it is clear that on their view what happens when I am consciously thinking about say the Middle East is that in some appropriate location in my overall cognitive operations there occurs a symbol (or set of symbols) that refers to the Middle East This symbol

f like the surrounding context in which it occurs is some neurophysiological state or some constellation of such states No one of course has at present any real

knowledge of the concrete nature of such symbols or their larger contexts but it f I will do no harm to follow Fodor in thinking of a mental blackboard on which

mental symbols are inscribed in appropriate ways Thus for me ro be consciously thinking about the Middle East is for me ro have the mental symbol that refers to

does not depend in a constitutive way on external causal rdations even though it may be causally produced by them Thus in that case too a knowledge of the persons internal physical and neurophysiological states alone should enable Mary to pick out One color rather than the other as the right one But it is even more obvious than in the original ase that this is not so

See eg Jerry Fodor Propositional Attitudes and Methodological Solipsism Considered as a Research Srrategy in Cognitive Science both reprinted in his 198 book Rtpwmll1rions (Cambridge MA MIT Press)

20 See for example Jerry Fodor chapter 4 Fodor has subsequently refined this view in various ways but none that affect issues being raised here

20 L Bonjour Against Materialism 21

the Middle East inscribed in the right way on this blackboard But the symbols reference to the Middle East to repeat depends not on its intrinsic physical or neurophysiological character alone but also on the relations in which it stands to other such symbols and directly or indirectly to the external world

Suppose now that Mary is studying my cognitive operations Suppose that she has somehow isolated what amounts to my mental blackboard and the various symbols written on it Obviously this will not in itself tell her what I am thinking about Even if she could somehow focus onthe specific symbol that refers to the Middle East and tell that it is functioning in a way that determines the object of my conscious thought (even though there is no reason to think that she could in fact do these things) she will not on this basis alone be able to tell what it is that this symbol in fact refers to Nor is there any plausibiliry to the idea that Mary could figure out the reference or meaning of the various mental symbols simply by examining their internal relations (0 each other21 Thus she will need once again to appeal to external causal relations ofvarious sorts

Bur how then am I able to be aware ofor understand from the inside what I am thinking about Once again I have no knowledge of those external relations (and would be very unlikely to be able to figure anything out from them even if I did) All that I plausibly have access to is the mental symbol or symbols and the surrounding system of states and this is apparently nor enough to determine the object of my thought

The only very obvious recourse here for the proponent of a language of thought is to construe my understanding or awareness of what I am thinking about disquotationally in relation to the language of thought When thinking about the Middle East I do so by using some mental symbol And when [ understand or am aware of what I am thinking about it might be suggested I in effect use that very same symbol what I am aware of is that I am thinking about the Middle East (whatever that is-that is whatever that symbol in fact refers (0) If the symbol in question did succeed in referring to the Middle East then this specification of what I am thinking about will refer to the Middle East as well and so will be correct But it is intuitively as obvious as anything could be that my awareness ofwhat I am thinking about normally involves more than this involves actually understanding (at some level of precision detail etc) what the Middle East is in a way that goes beyond merely repeating the same symbol Assuming for the moment that there really is a language of thought I understand my language of thought in a way parallel to the way in which I understand my own public language-and not in the merely disquotational way that could just as well be applied to a language of which I have no understanding at all

Here a proponent of the language of thought may want to reply that the difference in the public language case is merely that one language is a language

2 Such an idea has sometimes at least apparently been suggested For more discussion see 01)

19()8 book In Difmse ofPure Reason pp 174-80

that I successfully use-and that the same is true of my language of thought On this view the intuition that I understand what I am thinking about-or what I am talking about-in any stronger sense one that is not merely disquotational is merely an illusion But here again we have a view that it seems to me would appeal to no one who was not motivated by the conviction that materialism must be true

My conclusion is that the language of thought view has nothing useful to say about the most obvious sort of intentional content the intentional content that is involved in having something explicitly and consciously in mind Nor do I know of any other materialist account that does any better in this regard There is perhaps room for dispute about just how important conscious intentional content is in relation to the causation and explanation of behavior but no plausible way (0 deny that it genuinely exists Thus with respect to intentional content as with the case of qualitative content materialism seems to be utterly bankrupt as a general account of mental states and to be held merely as an article of faith

5 WHAT IS THE ALTERNATIVE

The last two sections serve merely to strengthen and deepen the fundamental objection to materialism already offered in section 2 consciousness genuinely exists materialism can oller no account that explains consciousness (or of the specific varieties thereof) or shows it [0 be merely material in character therefore (at least in the absence of any strong antecedent argument or presumption in favor of materialism) the indicated conclusion is that materialism is false There is more in heaven and earth than is dreamt of in materialist philosophy

But what do I mean by more Here as I see it there is very little that can be said in our current state of knowledge so thar the main result is that we have very little understanding ofconsciousness-or given the arguably central role of consciousness of mentaliry in general

In the first place there is no clear way in which the objections that I have raised against materialism support the classical substance dualist position Positing a separate mental substance that is characterized in almost entirely negative terms does nothing very obvious to explain consciousness in general or qualitative and intentional content in particular As far as I can see the main appeal of substance dualism is that the account of the supposed mental or spiritual substances is fur too vague and sketchy to provide the basis for any very clear argument that such substances could not be the locus of consciousness But this negative point hardly counts as an argument in favor of such a view

The obvious alternative is properry dualism the view that human persons and perhaps other kinds of animals have non-material or non-physical properties in addition to their physical ones with at least the main such properties being

10 L Bonjour Against Materialism 11

to offer some supplementary account of what material features give rise to

conscious experience But I know of no such account at least none with any real plausibiliry6

This difficulty with materialism in general and functionalism in particular has of course occasionally been recognized But it still seems to have had remarkably little impact on the widespread acceptance of materialist views I have no very good explanation to offer of this though part of the reason is perhaps the prevailing tendency to approach the philosophy of mind from a third-person neo-behaviorist perspective in which consciousness is largely or entirely ignored (But on this issue it is hard to distinguish the chickens from the eggs)

My basic case against materialism is complete at this point there is no good reason for any strong presumption in favor of materialism and the main materialist view fails to offer any real explanation of a central aspect of mental states namely their conscious character meaning that there is no good reason to think that it is correct as an account of such states But though this very simple argument seems to me entirely compelling I will elaborate it further in the next two sections by focusing on the two main specific kinds of mental states The version of the argument that applies to states with qualitative content is very familiar even though I think that its full force has still not been generally appreciated In contrast the application of essentially the same basic argument to conscious states with intentional content has received far less attention

3 THE PROBLEM OF QUALITATIVE CONSCIOUSNESS

MARY REDUX

Though functionalism fails to adequately account for consciousness of any SOrt

perhaps the most conspicuous aspect ofthis failure penains to qualitative content the sort of content involved in experiences of color and sound and of like pains and itches This point has been made in many ways but the most straightforward and compelling in my view is still the so-called argument initially suggested by Thomas Nagel in relation to the experiences of bats and later developed by Frank Jackson using his famous example of black-and-white Mary on which] will mainly focus here (As most will know

6 One possibiliry is the so-allIed higher order thought theory which holds that consciousne arises when one mental state is the object of a second higher order mental state I have no room here to consider this view in detaiL But the basic-and obvious-problem with such a view is that there is no reason why there could not be a hierarchy of sort even one with many more levels in which there was no consciousness involved at alL (For some elaboration see my contribution to

Bonjour and Sos (2003) pp 65-8) 7 See for example Colin McGinn (1989) David Chalmers (1995) and David Chalmers (1996)

Jackson has changed his mind about this argument and now rejects it though his reasons seem to me unpersuasive)

I will assume here that Jacksons original version of the saga of Mary is familiar enough to require only a brief summation Mary is a brilliant neurophysiologist who lives her entire life acquires her education and does all of her scientific work in a black-and-white environment using black-and-white books and blackshyand-white television for all of her learning and research In this way we may suppose she comes to have a complete knowledge of all the physical facts in neurophysiology and related fields together with their deductive consequences insofar as these are relevanr-thus arriving at as complete an understanding of human functioning as those sciences can provide In particular Mary knows the functional roles of all of the various neurophysiological states including those pertaining to visual perception by knowing their causal relations to sensory inputs behavioral outputs and other such states But despite all of this knowledge Mary apparently does nor know all that there is to know about human mental states for when she is released from her black-and-white environmenr and allowed to view the world normally she will by viewing objects like ripe tomatoes learn what it is like to see something red and analogous things about other qualitative experiences But then comments Jackson it is inescapable that her previous knowledge was incomplete But she had all the physical information Ergo there is more to have than that and Physicalism is falses

Despite the initial force of this rather simple argumenr materialists have nOt been persuaded and the literature comprising materialist responses to the Mary example is very large 9 One thing to say about these responses is that few if any of them are even claimed ro have any substanrial independent plausibility instead they are PUt forward in a way that takes for granted the sort of general presumption in favor of materialism and correlative burden of proof for antishymaterialist views that J have argued does not genuinely exist A full discussion of these responses is impossible here but there are some main themes that can be

dealt with in a general way 10 One of these is the suggestion that although something new when she leaves the black-and-white

room what she acquires is not a knowledge of a new foct (or facts) bur rather something else A second is the suggestion that what she does something like a new ability perhaps more specificall representational ability And if these rwo themes are

bull Frank Jackson (1982) p 130 see also Frank Jackson (1986) bull Many of these discussions are collected in Ludlow Nagasawa and

)0 For a useful taxonomy of the various possible materialist responses see Robert Van Gulick So Many Ways of Saying No to Mary in Ludlow Nagasawa and Stoljar (2004) pp 365-405 (This is of the places where the materialist discussion bears a striking similariry to scholastic theology one can easily imagine a complacent theist writing an artide entitled So Many Ways to Answer the Problem of Evil)

12 L Bonjour i Against Materialism 13

fresult is that there is nothing about the Mary example that is incompatible with Though we have made things vastly easier for Mary by focusing on twO cases the truth of materialism involving colors with which she is now familiar rather than asking her to figure

Particularly in light of rhe general materialist failure to provide an account out on the basis ofher overall physicalneurophysiologicalfunctional knowledge I of conscious experience I doubt very much whether any response of this son would seem even mildly convincing to anyone who was not already determined to adhere to materialism come what may But the first of these fWO themes does at least point to a kind of lacuna in Jacksons original account of the case if Mary learns new facts what exactly are they Indeed in his response to an early version of this suggestion Jackson is reduced to invoking the problem of other minds as a (very) indirect basis for thinking that factual knowledge of some son is involved II

It is however surprisingly easy to modifY the original case in a way that makes it utterly dear that there are facts that Mary does not know while she is in the black-and-white room and will learn when she emerges Suppose that while she is still in the otherwise black-and-white environment fWO color samples are brought in one a sample of a fairly bright green approximately the color of newly mown grass and the other a sample of a fairly bright red approximately the color of a fire engine Mary is allowed to view these samples and even to know that they are fWO of the colors that she has learned about in her black-and-white education She is not however told the standard names of these colors nor is she allowed to monitor her own neurophysiology as she views them

We now remind Mary of two specific ca~es that she has studied morougIllV and about which she knows all the physic~l One of these is a case where a person was second is a case where a person was lookmg at a tell Mary that one of these people had an experience predominar one of the colors with which she is now familiar and that the other person had an experience predominantly involving the other color but of course not which was which If we call the colors presented by the samples color A and color B Mary now knows that one of the two following pairs of claiflls is true

(l) The experience of freshly mown grass predominantly involves color A and the experience of a newly painted fire engine predominantly involve~ color B

(2) The experience of freshly mown grass predominantly involves color B and the experience ofa newly painted fire engine predominantly involves color A

But can she tell on the basis of her black-and-white knowledge together with her new familiarity with the two colors whether it is (1) or (2) which is true (Notice carefully that there is no apparent problem with her understanding of these claims)

l See Jackson (1986 294)

what color experiences in general are like it still seems quite dear for essentially the same reasons that were operative in the original case that she will have no more success with her much more limired task Just as there was nothing in the physical account that could tell her what an experience of red was like so there is still nothing in the physical account of the fire engine case that could definitively pick out the color of one of the samples as opposed to the other 12

And yet whichever of (1) and (2) is true states a foct (or facts) in as robust a sense as one could want-a fact that Mary will learn when she emerges from the black-and-white room and is allowed to view ordinary objects of various kinds

Moreover if there are abilities that result from experiencing the two colors in question Mary presumably can acquire them on the basis of the samples Consider for example Harmans suggestion l3 that what Mary acquires in the original case when she leaves the black-and-white room and sees red for the first time is a perceptual concept of red one that essentially involves being disposed to form perceptual representations involving it in the presence ofcausal stimulation of the right SOrt-sO that she cannot acquire that concept in the original version of the black-and-white room There is much that is questionable about the idea ofsuch a concept but if there is indeed such a thing then Mary in the new version of the case presumably can acquire it by viewing the red sample

more than one sample is for some reason required but it would be easy the new version of the case to allow for that) So we may

suppose Mary has the perceptual concept of red and the perceptual concept of green but she still cannot figure our from her physical knowledge which of these concepts is being employed by the people in the cases she has studied Yet this too is a fact and ifmaterialism isrrue an entirely physical fact So why cant she know it

Here as far as I can see there are only fWO possible moves for the materialist which are even marginally worth considering One is the suggestion that Mary already knows the facts in question as a part of her overall physical knowledge but that she knows them under a different guise or mode of presentation than that under which she will come to know them when she leaves the blackshyand-white room This idea can be developed in different ways and with enormous technical ingenuity But does it really have any serious plausibility Imagine that

J2 As Jackson emphasizes in Jackson (1986 295) it is nO enough for Mary (0 be able to conjecture or guess at the answerO this question For physicalism to be true the fact in question must actually be contained in her physical knowledge

U Gilbert Harman (1990) The Intrinsic Quality of Experience Philosophical Pmpectives 4 3J~52 at pp 44~5 Harman does not actually melllion the Maty case as sucn focusing instead on a person who is blind from birth but still learns all the physical and functional facts of color perception But he does cite Jackson (along with Nagel) as the source of tne objection he is discussing

14 15 L Bonjour

Mary in our modified version of the case having finally experienced real colors is eager to find out more about these intriguing features of the world about which she has been kept in ignorance She wants very much for example to know whether it is (I) or (2) that captures the relevant facts about cases of that sort-and is seriously frustrated about being kept in ignorance any longer Suppose that we respond to her frustration by informing her that she already knows the very facts that she is so eager to learn Surely she would not be satisfied How might she respond

I think we can imagine Mary saying something like this

You philosophers are really amazing The idea that I ar(atiy know the facts I am interested in-indeed all facts of that general kind-is simply preposterous I know all of the physical details but none of them tells me which of the properties I have just experienced on the basis of the samples is realized in each of the two cases If you suggest to me that there arent really novel properties but rather novel concepts or ways of representing or whatever then (while finding that suggestion itself pretry hard to swallow) I would still insist that which concept or way of representing is involved in each case is still something that my physical knowledge doesnt give me any clue about Perhaps as you say there is some clever or complicated way in which the things I want to know are related to the physical things I do know-maybe there is even some metaphysically necessary connection between them (assuming that it is kosher for materialists to believe in such things) Anything like that however just adds to the list of facts that my physical knowledge doesnt reveal to me I am a scientist and not a philosopher so Im not really sure which conception of a fact is the right one (All of the ones you suggest seem pretty weird) But there is undeniably something that I want to know-something that is true about the world-that cant be learned on the basis of all my physical knowledge And that means that the physical story isnt in fact the whole story

Not surprisingly I think that the response I have imagined for Mary is exactly right-that any way of understanding or individuating facts according to which some piece of Marys physical knowledge and either (1) or (2) above turn out to be formulations of the same fact is a conception of fact that is simply too intuitively implausible to be taken seriously

The other possible materialist response is to gram that Mary will learn new facts when she emerges from the black-and-white room in the modified version of the case but to insist that these are nonetheless still physical facts On this view what the case shows is that it is impossible for Mary to acquire complete physical knowledge in the black-and-white room One way to put it is to say that while she can learn all the objective physical facts there are still certain subjective physical facts l4 that she cant learn One can learn what it feels like subjectively to be an organism ofsuch-and-such a general physical description in such-and-such a specific physical state only by actually realizing that condition But that it feels

14 See Van Gulick (2004) for one version of this suggestion

Agaimt Materialism

a certain way or involves a certain sort of conscious experience is still on this view an entirely physical fact

I have to admit that I find it nearly impossible to take this response seriously The only argument for it seems to be an appeal to a background presumption in favor of materialism that is so strong as to make it allegedly reasonable to claim that any fact there is must be a material fact even ifwe cant see in any clear way how it could be a material fact Materialism as we have already seen offers no real account or explanation of consciousness and so also no reason for thinking that there is any subjective experience at all involved in being in a certain material state Thus to advance a view of this sort is in effect just to insist that no fact of any sort can be allowed to refute materialism and thus that any possibiliry of this sort must simply be absorbed into the materialist view however inexplicable in materialist terms it may be (It is not much of a stretch to imagine the materialist saying that we must first believe in order than we may understand)

Thus the modified version of the Mary case seems to present an objection to materialism in general (and functionalism in particular) that is about as conclusive as philosophical arguments ever get However exactly they should be characterized there are facts that Mary cannot know on the basis ofher complete physicalneurophysiologicalfunctional knowledge even when she is given the sort of limited experience needed to understand the claims in question and to acquire any abilities that might be relevant These facts do not seem to be material facts and there is no basis that is not utterly arbitrary and question-begging for supposing that they are Thus we have the strongest of reasons for holding that the materialist account of realiry is incomplete and hence that materialism is

false

4 THE PROBLEM OF CONSCIOUS INTENTIONAL

CONTENT

Qualia of the son involved in the Mary case are widely recognized to pose a serious problem for materialist views and it is not too hard to discern occasional misgivings in this respect under the faltade of materialist confidence But as already mentioned above it is widely assumed that materialism is in much better shape with regard to intentional mental states propositional attitudes and other states that involve aboutness I believe however that this is almost entirely an illusion-that the problems for materialism are just as serious in this area with consciousness being once again the central focus

Materialist accounts of intentional states tend to focus mainly on dispositional states such as beliefs and desires Given the central role of such states in explanations of behavior this is in some ways reasonable enough But such a focus tends to neglect or even ignore the existence of comcious intentional states-even though having conscious thoughts that P is surely one of the

16 L Bonjour

central things that having a dispositional belief that P disposes one to do A focus on belief in particular also has the unfortunate effect of making externalist accounts of intentional content seem more plausible than they possibly could if the emphasis were on conscious intentional states

For these reasons I will focus here explicitly on conscious thoughts As I sit writing this chapter a variety of conscious thoughts pass through my mind Many of these involve the assertion or endorsement of various propositions that materialism cannot account for consciousness that the trees outside my window are very bare that the weather looks cold and dank that the situation in the Middle East looks grim and so on Other thoughts are also propositional but in a way that does not involve assertion my conscious desire to get several pages written before lunch the hope that the stock market will continue to rise and so on It is doubtful that conscious thought must always be propositional in character but it will in any case simplify the issues to be discussed if we largely ignore the propositional aspect of these various thoughts and focus simply on their being conscious thoughts of or about various things or kinds of things materialism the trees the Middle East the stock market and so on

One crucial feature of such conscious thoughts is that when I have them I am in general consciously aware of or consciously understand or grasp what it is that I am thinking about (and also what I am thinking about it) When I think that the trees outside my window are bare I consciously understand that it is certain trees that I am thinking about (and along with this what sort of thing a tree is and which trees I have in mind) What exactly this conscious grasp of the object of thought involves varies from case to case and is sometimes not easy to precisely specify Moreover as will emerge it is something of which I think we presently have no real explanatory account of any substance But its existence is I submit completely undeniable Indeed being able in this way to consciously think about things to have them in mind is in many ways the most central and obvious feature of our mental lives

It is obvious that a persons conscious grasp of the object of their thought of what they are thinking about can vary on a number of dimensions it may be more or less precise more or less detailed more or less clear more or less complete But contrary to what is sometimes suggested it is rarely if ever merely disquotational in character Perhaps (though I doubt it) there are cases where a scientifically untutored person is thinking about eg electrons and where their sole grasp of what they are thinking is that it is what is referred to in their society or community by the word electron -so that what they are thinking about is in effect electrons (whatever they are) But this is surely not the ordinary situation when we think about various things 15

15 Notice that even a thought about what the relevant Ioci~tal ocp~rts mean by electrons would have to involve a non-disquotational element in the reference to those experts and also in the reference to the word to think about whatever it is that the societal experts mean by electrons

Against Materialism 17

Moreover the existence of conscious intentional content is perfectly compatshyible with the existence of an externalist dimension of thought content-though not with the view that all content is external If as in Putnams famous example a person is thinking about earthly water at a time prior to the discovery of its chemical composition there is no reason to deny that they are in a sense thinking about H 20 But in such a situation the aspect of being about H 20 will obviously not be part of their conscious internal grasp of what they are thinking about in the way that the more superficial aspects of water will 16 And in a somewhat parallel way the person in Burges famous example who thinks he has arthritis in his thigh and to whom our standard belief-ascription practices ascribe a belief about arthritis (where this is among other things a disease that only occurs in joints) obviously does not consciously grasp the disease that he is thinking about in a way that involves this specific feature of itY But it is nonetheless impossible to describe either example in a convincing way without presupposing that the people in question do have something consciously in mind a substance having the superficial properties of water in Putnams example and a disease having certain fairly specific features in Burges Thus while it is possible to dispute the relative importance ofconscious internal thought content and external thought content of which the subject is not conscious examples of this sort provide no basis at all for denying that conscious internal content exists

The issue I want to raise here is whether a materialist view can account for the sort of conscious intentional content just characterized Can it account for conscious thoughts being about various things in a way that can be grasped or understood by the person in question In a way the answer has already been given Since materialist views really take no account at all of consciousness they obviously offer no account of this particular aspect of it But investigating this narrower aspect of the issue can still help to deepen the basic objection to

materialism Here it will be useful to bring the brilliant neurophysiologist Mary briefly

back onto the scene even though the black-and-white aspect of her situation is no longer relevant Suppose that Mary studies me as a subject and comes to

have a complete knowledge of my physical and neurophysiological makeup as I am thinking these various thoughts Can she determine on that basis what I am consciously thinking about at a particular moment

One thing that seems utterly clear is that she could not do this merely on the basis of knowing my internal physical characteristics-as it is sometimes put

I knowing everything physical that happens inside my skin There is no reason t f

is not the same thing as thinking about whatever is meant by electrons by whatever is meant by the societal experts

I 16 See Hilary Putnam (1975a) (Cambridge Cambridge Universiry Press) pp 215-71 17 See Tyler Burge (1979)

19 18 L Bonjour

at all to think that the internal structure of my physical and neurophysiological states could somehow by itself determine that I am thinking about weather tather than about the Middle East or the stock market

A functionalist would no doubt say that it is no surprise that Mary could not do this In order to know the complete causal or functional role of my internal states Mary also needs to know about their external causal relations to various things And it might be suggested if Mary knows all of the external causal relations in which my various states stand she will in fact be able ro figure out what I am consciously thinking about at any particular time No doubt the details that pick out any particular object of thought will be very complicated but there is it might be claimed no reason to doubt that in principle she could do this

Here we have a piece of materialist doctrine that again has a status very similar to that of a claim of theology It is obvious that no one ha~ even the beginnings of an idea of how to actually carry out an investigation that would yield a result of this kind-that the only rea~on for thinking that this could be done is the overriding assumption for which we have found no cogent basis that materialism must be true Among a multitude of other difficulties Mary would have to be able to figure out the content of thoughts that are confused or inaccurate or thoughts about imaginary or fictional entities or supernatural entities It is to say the least very hard to see how she could do this on the basis of a knowledge of causal relations to more ordinary sorts of things

But the problem for materialism is in fact even worse than that For as already emphasized it is an undeniable fact about conscious intentional content that I am able for the most part to consciously understand or be aware of what I am thinking about from the inside Clearly I do not in general do this on the basis of external causal knowledge I do not have such knowledge and would not know what to do about it if I did All that I normally have any sort of direct access to if materialism is true is my own internal physical and physiological states and thus my conscious understanding of what I am thinking about at a particular moment must be somehow a feature or result of those internal states alone Causal relations to external things may help to produce the relevant features of the internal states in question but there is no apparent way in which such external relations can somehow be partly constitutive of the fact that my conscious thoughts are about various things in a way of which I can be immediately aware But if these internal states are sufficient to fix the object of my thought in a way that is accessible to my understanding or awareness then knowing about those internal states should be sufficient for Mary a~ well without any knowledge of the external causal relations And yet as we have already seen it seems obvious that this is not the ca~e18

18 Ir is worth noting that the same thing is really true in the ase of qualia as well A persons awareness of one color rather than another when he or she looks at newly mown grass obviously

Against Materialism

Thus we have the basis for an argument that is parallel to Jacksons original argument about qualia Mary knows all the relevant physical facts she is not able on the basis of this knowledge to know what I am consciously thinking about at a particular moment but what I am thinking about at that moment is as surely a fact about the world as anything else therefore complete physical knowledge is not complete knowledge and so materialism is false

One way to further elaborate this point is to consider how it applies to

what is perhaps the most widely held materialist view of intentional content the view popularized by Jerry Fodor and many others that intentional mental states employ an internal language a language of thought19 Fodor calls this view the representational theory of the mind though it might better be called the symbolic theory of the mind For the crucial feature of the view is that the language of thought like any language is composed of symbols items that do not stand for anything by virtue of their intrinsic properties but whose representative character depends instead on the relations in which they stand to other things-for Fodor the sorts of causal relations that are captured in the idea of a causal or functional role1o Just as the word dog could in principle have stood for anything (or nothing at all) and in fact stands for a kind ofanimal rather than something else only because of causal relations that arise from the way it is used so also the symbols in the language of thought stand for whatever they stand for only by virrue of analogous sorts of relations and not in virtue of their intrinsic physical and neurophysiological properties Their intentional character is thus extrinsic not intrinsic

Proponents of the language ofthought rarely have much to say about conscious thoughts of the sort that we are focusing on here But it is clear that on their view what happens when I am consciously thinking about say the Middle East is that in some appropriate location in my overall cognitive operations there occurs a symbol (or set of symbols) that refers to the Middle East This symbol

f like the surrounding context in which it occurs is some neurophysiological state or some constellation of such states No one of course has at present any real

knowledge of the concrete nature of such symbols or their larger contexts but it f I will do no harm to follow Fodor in thinking of a mental blackboard on which

mental symbols are inscribed in appropriate ways Thus for me ro be consciously thinking about the Middle East is for me ro have the mental symbol that refers to

does not depend in a constitutive way on external causal rdations even though it may be causally produced by them Thus in that case too a knowledge of the persons internal physical and neurophysiological states alone should enable Mary to pick out One color rather than the other as the right one But it is even more obvious than in the original ase that this is not so

See eg Jerry Fodor Propositional Attitudes and Methodological Solipsism Considered as a Research Srrategy in Cognitive Science both reprinted in his 198 book Rtpwmll1rions (Cambridge MA MIT Press)

20 See for example Jerry Fodor chapter 4 Fodor has subsequently refined this view in various ways but none that affect issues being raised here

20 L Bonjour Against Materialism 21

the Middle East inscribed in the right way on this blackboard But the symbols reference to the Middle East to repeat depends not on its intrinsic physical or neurophysiological character alone but also on the relations in which it stands to other such symbols and directly or indirectly to the external world

Suppose now that Mary is studying my cognitive operations Suppose that she has somehow isolated what amounts to my mental blackboard and the various symbols written on it Obviously this will not in itself tell her what I am thinking about Even if she could somehow focus onthe specific symbol that refers to the Middle East and tell that it is functioning in a way that determines the object of my conscious thought (even though there is no reason to think that she could in fact do these things) she will not on this basis alone be able to tell what it is that this symbol in fact refers to Nor is there any plausibiliry to the idea that Mary could figure out the reference or meaning of the various mental symbols simply by examining their internal relations (0 each other21 Thus she will need once again to appeal to external causal relations ofvarious sorts

Bur how then am I able to be aware ofor understand from the inside what I am thinking about Once again I have no knowledge of those external relations (and would be very unlikely to be able to figure anything out from them even if I did) All that I plausibly have access to is the mental symbol or symbols and the surrounding system of states and this is apparently nor enough to determine the object of my thought

The only very obvious recourse here for the proponent of a language of thought is to construe my understanding or awareness of what I am thinking about disquotationally in relation to the language of thought When thinking about the Middle East I do so by using some mental symbol And when [ understand or am aware of what I am thinking about it might be suggested I in effect use that very same symbol what I am aware of is that I am thinking about the Middle East (whatever that is-that is whatever that symbol in fact refers (0) If the symbol in question did succeed in referring to the Middle East then this specification of what I am thinking about will refer to the Middle East as well and so will be correct But it is intuitively as obvious as anything could be that my awareness ofwhat I am thinking about normally involves more than this involves actually understanding (at some level of precision detail etc) what the Middle East is in a way that goes beyond merely repeating the same symbol Assuming for the moment that there really is a language of thought I understand my language of thought in a way parallel to the way in which I understand my own public language-and not in the merely disquotational way that could just as well be applied to a language of which I have no understanding at all

Here a proponent of the language of thought may want to reply that the difference in the public language case is merely that one language is a language

2 Such an idea has sometimes at least apparently been suggested For more discussion see 01)

19()8 book In Difmse ofPure Reason pp 174-80

that I successfully use-and that the same is true of my language of thought On this view the intuition that I understand what I am thinking about-or what I am talking about-in any stronger sense one that is not merely disquotational is merely an illusion But here again we have a view that it seems to me would appeal to no one who was not motivated by the conviction that materialism must be true

My conclusion is that the language of thought view has nothing useful to say about the most obvious sort of intentional content the intentional content that is involved in having something explicitly and consciously in mind Nor do I know of any other materialist account that does any better in this regard There is perhaps room for dispute about just how important conscious intentional content is in relation to the causation and explanation of behavior but no plausible way (0 deny that it genuinely exists Thus with respect to intentional content as with the case of qualitative content materialism seems to be utterly bankrupt as a general account of mental states and to be held merely as an article of faith

5 WHAT IS THE ALTERNATIVE

The last two sections serve merely to strengthen and deepen the fundamental objection to materialism already offered in section 2 consciousness genuinely exists materialism can oller no account that explains consciousness (or of the specific varieties thereof) or shows it [0 be merely material in character therefore (at least in the absence of any strong antecedent argument or presumption in favor of materialism) the indicated conclusion is that materialism is false There is more in heaven and earth than is dreamt of in materialist philosophy

But what do I mean by more Here as I see it there is very little that can be said in our current state of knowledge so thar the main result is that we have very little understanding ofconsciousness-or given the arguably central role of consciousness of mentaliry in general

In the first place there is no clear way in which the objections that I have raised against materialism support the classical substance dualist position Positing a separate mental substance that is characterized in almost entirely negative terms does nothing very obvious to explain consciousness in general or qualitative and intentional content in particular As far as I can see the main appeal of substance dualism is that the account of the supposed mental or spiritual substances is fur too vague and sketchy to provide the basis for any very clear argument that such substances could not be the locus of consciousness But this negative point hardly counts as an argument in favor of such a view

The obvious alternative is properry dualism the view that human persons and perhaps other kinds of animals have non-material or non-physical properties in addition to their physical ones with at least the main such properties being

12 L Bonjour i Against Materialism 13

fresult is that there is nothing about the Mary example that is incompatible with Though we have made things vastly easier for Mary by focusing on twO cases the truth of materialism involving colors with which she is now familiar rather than asking her to figure

Particularly in light of rhe general materialist failure to provide an account out on the basis ofher overall physicalneurophysiologicalfunctional knowledge I of conscious experience I doubt very much whether any response of this son would seem even mildly convincing to anyone who was not already determined to adhere to materialism come what may But the first of these fWO themes does at least point to a kind of lacuna in Jacksons original account of the case if Mary learns new facts what exactly are they Indeed in his response to an early version of this suggestion Jackson is reduced to invoking the problem of other minds as a (very) indirect basis for thinking that factual knowledge of some son is involved II

It is however surprisingly easy to modifY the original case in a way that makes it utterly dear that there are facts that Mary does not know while she is in the black-and-white room and will learn when she emerges Suppose that while she is still in the otherwise black-and-white environment fWO color samples are brought in one a sample of a fairly bright green approximately the color of newly mown grass and the other a sample of a fairly bright red approximately the color of a fire engine Mary is allowed to view these samples and even to know that they are fWO of the colors that she has learned about in her black-and-white education She is not however told the standard names of these colors nor is she allowed to monitor her own neurophysiology as she views them

We now remind Mary of two specific ca~es that she has studied morougIllV and about which she knows all the physic~l One of these is a case where a person was second is a case where a person was lookmg at a tell Mary that one of these people had an experience predominar one of the colors with which she is now familiar and that the other person had an experience predominantly involving the other color but of course not which was which If we call the colors presented by the samples color A and color B Mary now knows that one of the two following pairs of claiflls is true

(l) The experience of freshly mown grass predominantly involves color A and the experience of a newly painted fire engine predominantly involve~ color B

(2) The experience of freshly mown grass predominantly involves color B and the experience ofa newly painted fire engine predominantly involves color A

But can she tell on the basis of her black-and-white knowledge together with her new familiarity with the two colors whether it is (1) or (2) which is true (Notice carefully that there is no apparent problem with her understanding of these claims)

l See Jackson (1986 294)

what color experiences in general are like it still seems quite dear for essentially the same reasons that were operative in the original case that she will have no more success with her much more limired task Just as there was nothing in the physical account that could tell her what an experience of red was like so there is still nothing in the physical account of the fire engine case that could definitively pick out the color of one of the samples as opposed to the other 12

And yet whichever of (1) and (2) is true states a foct (or facts) in as robust a sense as one could want-a fact that Mary will learn when she emerges from the black-and-white room and is allowed to view ordinary objects of various kinds

Moreover if there are abilities that result from experiencing the two colors in question Mary presumably can acquire them on the basis of the samples Consider for example Harmans suggestion l3 that what Mary acquires in the original case when she leaves the black-and-white room and sees red for the first time is a perceptual concept of red one that essentially involves being disposed to form perceptual representations involving it in the presence ofcausal stimulation of the right SOrt-sO that she cannot acquire that concept in the original version of the black-and-white room There is much that is questionable about the idea ofsuch a concept but if there is indeed such a thing then Mary in the new version of the case presumably can acquire it by viewing the red sample

more than one sample is for some reason required but it would be easy the new version of the case to allow for that) So we may

suppose Mary has the perceptual concept of red and the perceptual concept of green but she still cannot figure our from her physical knowledge which of these concepts is being employed by the people in the cases she has studied Yet this too is a fact and ifmaterialism isrrue an entirely physical fact So why cant she know it

Here as far as I can see there are only fWO possible moves for the materialist which are even marginally worth considering One is the suggestion that Mary already knows the facts in question as a part of her overall physical knowledge but that she knows them under a different guise or mode of presentation than that under which she will come to know them when she leaves the blackshyand-white room This idea can be developed in different ways and with enormous technical ingenuity But does it really have any serious plausibility Imagine that

J2 As Jackson emphasizes in Jackson (1986 295) it is nO enough for Mary (0 be able to conjecture or guess at the answerO this question For physicalism to be true the fact in question must actually be contained in her physical knowledge

U Gilbert Harman (1990) The Intrinsic Quality of Experience Philosophical Pmpectives 4 3J~52 at pp 44~5 Harman does not actually melllion the Maty case as sucn focusing instead on a person who is blind from birth but still learns all the physical and functional facts of color perception But he does cite Jackson (along with Nagel) as the source of tne objection he is discussing

14 15 L Bonjour

Mary in our modified version of the case having finally experienced real colors is eager to find out more about these intriguing features of the world about which she has been kept in ignorance She wants very much for example to know whether it is (I) or (2) that captures the relevant facts about cases of that sort-and is seriously frustrated about being kept in ignorance any longer Suppose that we respond to her frustration by informing her that she already knows the very facts that she is so eager to learn Surely she would not be satisfied How might she respond

I think we can imagine Mary saying something like this

You philosophers are really amazing The idea that I ar(atiy know the facts I am interested in-indeed all facts of that general kind-is simply preposterous I know all of the physical details but none of them tells me which of the properties I have just experienced on the basis of the samples is realized in each of the two cases If you suggest to me that there arent really novel properties but rather novel concepts or ways of representing or whatever then (while finding that suggestion itself pretry hard to swallow) I would still insist that which concept or way of representing is involved in each case is still something that my physical knowledge doesnt give me any clue about Perhaps as you say there is some clever or complicated way in which the things I want to know are related to the physical things I do know-maybe there is even some metaphysically necessary connection between them (assuming that it is kosher for materialists to believe in such things) Anything like that however just adds to the list of facts that my physical knowledge doesnt reveal to me I am a scientist and not a philosopher so Im not really sure which conception of a fact is the right one (All of the ones you suggest seem pretty weird) But there is undeniably something that I want to know-something that is true about the world-that cant be learned on the basis of all my physical knowledge And that means that the physical story isnt in fact the whole story

Not surprisingly I think that the response I have imagined for Mary is exactly right-that any way of understanding or individuating facts according to which some piece of Marys physical knowledge and either (1) or (2) above turn out to be formulations of the same fact is a conception of fact that is simply too intuitively implausible to be taken seriously

The other possible materialist response is to gram that Mary will learn new facts when she emerges from the black-and-white room in the modified version of the case but to insist that these are nonetheless still physical facts On this view what the case shows is that it is impossible for Mary to acquire complete physical knowledge in the black-and-white room One way to put it is to say that while she can learn all the objective physical facts there are still certain subjective physical facts l4 that she cant learn One can learn what it feels like subjectively to be an organism ofsuch-and-such a general physical description in such-and-such a specific physical state only by actually realizing that condition But that it feels

14 See Van Gulick (2004) for one version of this suggestion

Agaimt Materialism

a certain way or involves a certain sort of conscious experience is still on this view an entirely physical fact

I have to admit that I find it nearly impossible to take this response seriously The only argument for it seems to be an appeal to a background presumption in favor of materialism that is so strong as to make it allegedly reasonable to claim that any fact there is must be a material fact even ifwe cant see in any clear way how it could be a material fact Materialism as we have already seen offers no real account or explanation of consciousness and so also no reason for thinking that there is any subjective experience at all involved in being in a certain material state Thus to advance a view of this sort is in effect just to insist that no fact of any sort can be allowed to refute materialism and thus that any possibiliry of this sort must simply be absorbed into the materialist view however inexplicable in materialist terms it may be (It is not much of a stretch to imagine the materialist saying that we must first believe in order than we may understand)

Thus the modified version of the Mary case seems to present an objection to materialism in general (and functionalism in particular) that is about as conclusive as philosophical arguments ever get However exactly they should be characterized there are facts that Mary cannot know on the basis ofher complete physicalneurophysiologicalfunctional knowledge even when she is given the sort of limited experience needed to understand the claims in question and to acquire any abilities that might be relevant These facts do not seem to be material facts and there is no basis that is not utterly arbitrary and question-begging for supposing that they are Thus we have the strongest of reasons for holding that the materialist account of realiry is incomplete and hence that materialism is

false

4 THE PROBLEM OF CONSCIOUS INTENTIONAL

CONTENT

Qualia of the son involved in the Mary case are widely recognized to pose a serious problem for materialist views and it is not too hard to discern occasional misgivings in this respect under the faltade of materialist confidence But as already mentioned above it is widely assumed that materialism is in much better shape with regard to intentional mental states propositional attitudes and other states that involve aboutness I believe however that this is almost entirely an illusion-that the problems for materialism are just as serious in this area with consciousness being once again the central focus

Materialist accounts of intentional states tend to focus mainly on dispositional states such as beliefs and desires Given the central role of such states in explanations of behavior this is in some ways reasonable enough But such a focus tends to neglect or even ignore the existence of comcious intentional states-even though having conscious thoughts that P is surely one of the

16 L Bonjour

central things that having a dispositional belief that P disposes one to do A focus on belief in particular also has the unfortunate effect of making externalist accounts of intentional content seem more plausible than they possibly could if the emphasis were on conscious intentional states

For these reasons I will focus here explicitly on conscious thoughts As I sit writing this chapter a variety of conscious thoughts pass through my mind Many of these involve the assertion or endorsement of various propositions that materialism cannot account for consciousness that the trees outside my window are very bare that the weather looks cold and dank that the situation in the Middle East looks grim and so on Other thoughts are also propositional but in a way that does not involve assertion my conscious desire to get several pages written before lunch the hope that the stock market will continue to rise and so on It is doubtful that conscious thought must always be propositional in character but it will in any case simplify the issues to be discussed if we largely ignore the propositional aspect of these various thoughts and focus simply on their being conscious thoughts of or about various things or kinds of things materialism the trees the Middle East the stock market and so on

One crucial feature of such conscious thoughts is that when I have them I am in general consciously aware of or consciously understand or grasp what it is that I am thinking about (and also what I am thinking about it) When I think that the trees outside my window are bare I consciously understand that it is certain trees that I am thinking about (and along with this what sort of thing a tree is and which trees I have in mind) What exactly this conscious grasp of the object of thought involves varies from case to case and is sometimes not easy to precisely specify Moreover as will emerge it is something of which I think we presently have no real explanatory account of any substance But its existence is I submit completely undeniable Indeed being able in this way to consciously think about things to have them in mind is in many ways the most central and obvious feature of our mental lives

It is obvious that a persons conscious grasp of the object of their thought of what they are thinking about can vary on a number of dimensions it may be more or less precise more or less detailed more or less clear more or less complete But contrary to what is sometimes suggested it is rarely if ever merely disquotational in character Perhaps (though I doubt it) there are cases where a scientifically untutored person is thinking about eg electrons and where their sole grasp of what they are thinking is that it is what is referred to in their society or community by the word electron -so that what they are thinking about is in effect electrons (whatever they are) But this is surely not the ordinary situation when we think about various things 15

15 Notice that even a thought about what the relevant Ioci~tal ocp~rts mean by electrons would have to involve a non-disquotational element in the reference to those experts and also in the reference to the word to think about whatever it is that the societal experts mean by electrons

Against Materialism 17

Moreover the existence of conscious intentional content is perfectly compatshyible with the existence of an externalist dimension of thought content-though not with the view that all content is external If as in Putnams famous example a person is thinking about earthly water at a time prior to the discovery of its chemical composition there is no reason to deny that they are in a sense thinking about H 20 But in such a situation the aspect of being about H 20 will obviously not be part of their conscious internal grasp of what they are thinking about in the way that the more superficial aspects of water will 16 And in a somewhat parallel way the person in Burges famous example who thinks he has arthritis in his thigh and to whom our standard belief-ascription practices ascribe a belief about arthritis (where this is among other things a disease that only occurs in joints) obviously does not consciously grasp the disease that he is thinking about in a way that involves this specific feature of itY But it is nonetheless impossible to describe either example in a convincing way without presupposing that the people in question do have something consciously in mind a substance having the superficial properties of water in Putnams example and a disease having certain fairly specific features in Burges Thus while it is possible to dispute the relative importance ofconscious internal thought content and external thought content of which the subject is not conscious examples of this sort provide no basis at all for denying that conscious internal content exists

The issue I want to raise here is whether a materialist view can account for the sort of conscious intentional content just characterized Can it account for conscious thoughts being about various things in a way that can be grasped or understood by the person in question In a way the answer has already been given Since materialist views really take no account at all of consciousness they obviously offer no account of this particular aspect of it But investigating this narrower aspect of the issue can still help to deepen the basic objection to

materialism Here it will be useful to bring the brilliant neurophysiologist Mary briefly

back onto the scene even though the black-and-white aspect of her situation is no longer relevant Suppose that Mary studies me as a subject and comes to

have a complete knowledge of my physical and neurophysiological makeup as I am thinking these various thoughts Can she determine on that basis what I am consciously thinking about at a particular moment

One thing that seems utterly clear is that she could not do this merely on the basis of knowing my internal physical characteristics-as it is sometimes put

I knowing everything physical that happens inside my skin There is no reason t f

is not the same thing as thinking about whatever is meant by electrons by whatever is meant by the societal experts

I 16 See Hilary Putnam (1975a) (Cambridge Cambridge Universiry Press) pp 215-71 17 See Tyler Burge (1979)

19 18 L Bonjour

at all to think that the internal structure of my physical and neurophysiological states could somehow by itself determine that I am thinking about weather tather than about the Middle East or the stock market

A functionalist would no doubt say that it is no surprise that Mary could not do this In order to know the complete causal or functional role of my internal states Mary also needs to know about their external causal relations to various things And it might be suggested if Mary knows all of the external causal relations in which my various states stand she will in fact be able ro figure out what I am consciously thinking about at any particular time No doubt the details that pick out any particular object of thought will be very complicated but there is it might be claimed no reason to doubt that in principle she could do this

Here we have a piece of materialist doctrine that again has a status very similar to that of a claim of theology It is obvious that no one ha~ even the beginnings of an idea of how to actually carry out an investigation that would yield a result of this kind-that the only rea~on for thinking that this could be done is the overriding assumption for which we have found no cogent basis that materialism must be true Among a multitude of other difficulties Mary would have to be able to figure out the content of thoughts that are confused or inaccurate or thoughts about imaginary or fictional entities or supernatural entities It is to say the least very hard to see how she could do this on the basis of a knowledge of causal relations to more ordinary sorts of things

But the problem for materialism is in fact even worse than that For as already emphasized it is an undeniable fact about conscious intentional content that I am able for the most part to consciously understand or be aware of what I am thinking about from the inside Clearly I do not in general do this on the basis of external causal knowledge I do not have such knowledge and would not know what to do about it if I did All that I normally have any sort of direct access to if materialism is true is my own internal physical and physiological states and thus my conscious understanding of what I am thinking about at a particular moment must be somehow a feature or result of those internal states alone Causal relations to external things may help to produce the relevant features of the internal states in question but there is no apparent way in which such external relations can somehow be partly constitutive of the fact that my conscious thoughts are about various things in a way of which I can be immediately aware But if these internal states are sufficient to fix the object of my thought in a way that is accessible to my understanding or awareness then knowing about those internal states should be sufficient for Mary a~ well without any knowledge of the external causal relations And yet as we have already seen it seems obvious that this is not the ca~e18

18 Ir is worth noting that the same thing is really true in the ase of qualia as well A persons awareness of one color rather than another when he or she looks at newly mown grass obviously

Against Materialism

Thus we have the basis for an argument that is parallel to Jacksons original argument about qualia Mary knows all the relevant physical facts she is not able on the basis of this knowledge to know what I am consciously thinking about at a particular moment but what I am thinking about at that moment is as surely a fact about the world as anything else therefore complete physical knowledge is not complete knowledge and so materialism is false

One way to further elaborate this point is to consider how it applies to

what is perhaps the most widely held materialist view of intentional content the view popularized by Jerry Fodor and many others that intentional mental states employ an internal language a language of thought19 Fodor calls this view the representational theory of the mind though it might better be called the symbolic theory of the mind For the crucial feature of the view is that the language of thought like any language is composed of symbols items that do not stand for anything by virtue of their intrinsic properties but whose representative character depends instead on the relations in which they stand to other things-for Fodor the sorts of causal relations that are captured in the idea of a causal or functional role1o Just as the word dog could in principle have stood for anything (or nothing at all) and in fact stands for a kind ofanimal rather than something else only because of causal relations that arise from the way it is used so also the symbols in the language of thought stand for whatever they stand for only by virrue of analogous sorts of relations and not in virtue of their intrinsic physical and neurophysiological properties Their intentional character is thus extrinsic not intrinsic

Proponents of the language ofthought rarely have much to say about conscious thoughts of the sort that we are focusing on here But it is clear that on their view what happens when I am consciously thinking about say the Middle East is that in some appropriate location in my overall cognitive operations there occurs a symbol (or set of symbols) that refers to the Middle East This symbol

f like the surrounding context in which it occurs is some neurophysiological state or some constellation of such states No one of course has at present any real

knowledge of the concrete nature of such symbols or their larger contexts but it f I will do no harm to follow Fodor in thinking of a mental blackboard on which

mental symbols are inscribed in appropriate ways Thus for me ro be consciously thinking about the Middle East is for me ro have the mental symbol that refers to

does not depend in a constitutive way on external causal rdations even though it may be causally produced by them Thus in that case too a knowledge of the persons internal physical and neurophysiological states alone should enable Mary to pick out One color rather than the other as the right one But it is even more obvious than in the original ase that this is not so

See eg Jerry Fodor Propositional Attitudes and Methodological Solipsism Considered as a Research Srrategy in Cognitive Science both reprinted in his 198 book Rtpwmll1rions (Cambridge MA MIT Press)

20 See for example Jerry Fodor chapter 4 Fodor has subsequently refined this view in various ways but none that affect issues being raised here

20 L Bonjour Against Materialism 21

the Middle East inscribed in the right way on this blackboard But the symbols reference to the Middle East to repeat depends not on its intrinsic physical or neurophysiological character alone but also on the relations in which it stands to other such symbols and directly or indirectly to the external world

Suppose now that Mary is studying my cognitive operations Suppose that she has somehow isolated what amounts to my mental blackboard and the various symbols written on it Obviously this will not in itself tell her what I am thinking about Even if she could somehow focus onthe specific symbol that refers to the Middle East and tell that it is functioning in a way that determines the object of my conscious thought (even though there is no reason to think that she could in fact do these things) she will not on this basis alone be able to tell what it is that this symbol in fact refers to Nor is there any plausibiliry to the idea that Mary could figure out the reference or meaning of the various mental symbols simply by examining their internal relations (0 each other21 Thus she will need once again to appeal to external causal relations ofvarious sorts

Bur how then am I able to be aware ofor understand from the inside what I am thinking about Once again I have no knowledge of those external relations (and would be very unlikely to be able to figure anything out from them even if I did) All that I plausibly have access to is the mental symbol or symbols and the surrounding system of states and this is apparently nor enough to determine the object of my thought

The only very obvious recourse here for the proponent of a language of thought is to construe my understanding or awareness of what I am thinking about disquotationally in relation to the language of thought When thinking about the Middle East I do so by using some mental symbol And when [ understand or am aware of what I am thinking about it might be suggested I in effect use that very same symbol what I am aware of is that I am thinking about the Middle East (whatever that is-that is whatever that symbol in fact refers (0) If the symbol in question did succeed in referring to the Middle East then this specification of what I am thinking about will refer to the Middle East as well and so will be correct But it is intuitively as obvious as anything could be that my awareness ofwhat I am thinking about normally involves more than this involves actually understanding (at some level of precision detail etc) what the Middle East is in a way that goes beyond merely repeating the same symbol Assuming for the moment that there really is a language of thought I understand my language of thought in a way parallel to the way in which I understand my own public language-and not in the merely disquotational way that could just as well be applied to a language of which I have no understanding at all

Here a proponent of the language of thought may want to reply that the difference in the public language case is merely that one language is a language

2 Such an idea has sometimes at least apparently been suggested For more discussion see 01)

19()8 book In Difmse ofPure Reason pp 174-80

that I successfully use-and that the same is true of my language of thought On this view the intuition that I understand what I am thinking about-or what I am talking about-in any stronger sense one that is not merely disquotational is merely an illusion But here again we have a view that it seems to me would appeal to no one who was not motivated by the conviction that materialism must be true

My conclusion is that the language of thought view has nothing useful to say about the most obvious sort of intentional content the intentional content that is involved in having something explicitly and consciously in mind Nor do I know of any other materialist account that does any better in this regard There is perhaps room for dispute about just how important conscious intentional content is in relation to the causation and explanation of behavior but no plausible way (0 deny that it genuinely exists Thus with respect to intentional content as with the case of qualitative content materialism seems to be utterly bankrupt as a general account of mental states and to be held merely as an article of faith

5 WHAT IS THE ALTERNATIVE

The last two sections serve merely to strengthen and deepen the fundamental objection to materialism already offered in section 2 consciousness genuinely exists materialism can oller no account that explains consciousness (or of the specific varieties thereof) or shows it [0 be merely material in character therefore (at least in the absence of any strong antecedent argument or presumption in favor of materialism) the indicated conclusion is that materialism is false There is more in heaven and earth than is dreamt of in materialist philosophy

But what do I mean by more Here as I see it there is very little that can be said in our current state of knowledge so thar the main result is that we have very little understanding ofconsciousness-or given the arguably central role of consciousness of mentaliry in general

In the first place there is no clear way in which the objections that I have raised against materialism support the classical substance dualist position Positing a separate mental substance that is characterized in almost entirely negative terms does nothing very obvious to explain consciousness in general or qualitative and intentional content in particular As far as I can see the main appeal of substance dualism is that the account of the supposed mental or spiritual substances is fur too vague and sketchy to provide the basis for any very clear argument that such substances could not be the locus of consciousness But this negative point hardly counts as an argument in favor of such a view

The obvious alternative is properry dualism the view that human persons and perhaps other kinds of animals have non-material or non-physical properties in addition to their physical ones with at least the main such properties being

14 15 L Bonjour

Mary in our modified version of the case having finally experienced real colors is eager to find out more about these intriguing features of the world about which she has been kept in ignorance She wants very much for example to know whether it is (I) or (2) that captures the relevant facts about cases of that sort-and is seriously frustrated about being kept in ignorance any longer Suppose that we respond to her frustration by informing her that she already knows the very facts that she is so eager to learn Surely she would not be satisfied How might she respond

I think we can imagine Mary saying something like this

You philosophers are really amazing The idea that I ar(atiy know the facts I am interested in-indeed all facts of that general kind-is simply preposterous I know all of the physical details but none of them tells me which of the properties I have just experienced on the basis of the samples is realized in each of the two cases If you suggest to me that there arent really novel properties but rather novel concepts or ways of representing or whatever then (while finding that suggestion itself pretry hard to swallow) I would still insist that which concept or way of representing is involved in each case is still something that my physical knowledge doesnt give me any clue about Perhaps as you say there is some clever or complicated way in which the things I want to know are related to the physical things I do know-maybe there is even some metaphysically necessary connection between them (assuming that it is kosher for materialists to believe in such things) Anything like that however just adds to the list of facts that my physical knowledge doesnt reveal to me I am a scientist and not a philosopher so Im not really sure which conception of a fact is the right one (All of the ones you suggest seem pretty weird) But there is undeniably something that I want to know-something that is true about the world-that cant be learned on the basis of all my physical knowledge And that means that the physical story isnt in fact the whole story

Not surprisingly I think that the response I have imagined for Mary is exactly right-that any way of understanding or individuating facts according to which some piece of Marys physical knowledge and either (1) or (2) above turn out to be formulations of the same fact is a conception of fact that is simply too intuitively implausible to be taken seriously

The other possible materialist response is to gram that Mary will learn new facts when she emerges from the black-and-white room in the modified version of the case but to insist that these are nonetheless still physical facts On this view what the case shows is that it is impossible for Mary to acquire complete physical knowledge in the black-and-white room One way to put it is to say that while she can learn all the objective physical facts there are still certain subjective physical facts l4 that she cant learn One can learn what it feels like subjectively to be an organism ofsuch-and-such a general physical description in such-and-such a specific physical state only by actually realizing that condition But that it feels

14 See Van Gulick (2004) for one version of this suggestion

Agaimt Materialism

a certain way or involves a certain sort of conscious experience is still on this view an entirely physical fact

I have to admit that I find it nearly impossible to take this response seriously The only argument for it seems to be an appeal to a background presumption in favor of materialism that is so strong as to make it allegedly reasonable to claim that any fact there is must be a material fact even ifwe cant see in any clear way how it could be a material fact Materialism as we have already seen offers no real account or explanation of consciousness and so also no reason for thinking that there is any subjective experience at all involved in being in a certain material state Thus to advance a view of this sort is in effect just to insist that no fact of any sort can be allowed to refute materialism and thus that any possibiliry of this sort must simply be absorbed into the materialist view however inexplicable in materialist terms it may be (It is not much of a stretch to imagine the materialist saying that we must first believe in order than we may understand)

Thus the modified version of the Mary case seems to present an objection to materialism in general (and functionalism in particular) that is about as conclusive as philosophical arguments ever get However exactly they should be characterized there are facts that Mary cannot know on the basis ofher complete physicalneurophysiologicalfunctional knowledge even when she is given the sort of limited experience needed to understand the claims in question and to acquire any abilities that might be relevant These facts do not seem to be material facts and there is no basis that is not utterly arbitrary and question-begging for supposing that they are Thus we have the strongest of reasons for holding that the materialist account of realiry is incomplete and hence that materialism is

false

4 THE PROBLEM OF CONSCIOUS INTENTIONAL

CONTENT

Qualia of the son involved in the Mary case are widely recognized to pose a serious problem for materialist views and it is not too hard to discern occasional misgivings in this respect under the faltade of materialist confidence But as already mentioned above it is widely assumed that materialism is in much better shape with regard to intentional mental states propositional attitudes and other states that involve aboutness I believe however that this is almost entirely an illusion-that the problems for materialism are just as serious in this area with consciousness being once again the central focus

Materialist accounts of intentional states tend to focus mainly on dispositional states such as beliefs and desires Given the central role of such states in explanations of behavior this is in some ways reasonable enough But such a focus tends to neglect or even ignore the existence of comcious intentional states-even though having conscious thoughts that P is surely one of the

16 L Bonjour

central things that having a dispositional belief that P disposes one to do A focus on belief in particular also has the unfortunate effect of making externalist accounts of intentional content seem more plausible than they possibly could if the emphasis were on conscious intentional states

For these reasons I will focus here explicitly on conscious thoughts As I sit writing this chapter a variety of conscious thoughts pass through my mind Many of these involve the assertion or endorsement of various propositions that materialism cannot account for consciousness that the trees outside my window are very bare that the weather looks cold and dank that the situation in the Middle East looks grim and so on Other thoughts are also propositional but in a way that does not involve assertion my conscious desire to get several pages written before lunch the hope that the stock market will continue to rise and so on It is doubtful that conscious thought must always be propositional in character but it will in any case simplify the issues to be discussed if we largely ignore the propositional aspect of these various thoughts and focus simply on their being conscious thoughts of or about various things or kinds of things materialism the trees the Middle East the stock market and so on

One crucial feature of such conscious thoughts is that when I have them I am in general consciously aware of or consciously understand or grasp what it is that I am thinking about (and also what I am thinking about it) When I think that the trees outside my window are bare I consciously understand that it is certain trees that I am thinking about (and along with this what sort of thing a tree is and which trees I have in mind) What exactly this conscious grasp of the object of thought involves varies from case to case and is sometimes not easy to precisely specify Moreover as will emerge it is something of which I think we presently have no real explanatory account of any substance But its existence is I submit completely undeniable Indeed being able in this way to consciously think about things to have them in mind is in many ways the most central and obvious feature of our mental lives

It is obvious that a persons conscious grasp of the object of their thought of what they are thinking about can vary on a number of dimensions it may be more or less precise more or less detailed more or less clear more or less complete But contrary to what is sometimes suggested it is rarely if ever merely disquotational in character Perhaps (though I doubt it) there are cases where a scientifically untutored person is thinking about eg electrons and where their sole grasp of what they are thinking is that it is what is referred to in their society or community by the word electron -so that what they are thinking about is in effect electrons (whatever they are) But this is surely not the ordinary situation when we think about various things 15

15 Notice that even a thought about what the relevant Ioci~tal ocp~rts mean by electrons would have to involve a non-disquotational element in the reference to those experts and also in the reference to the word to think about whatever it is that the societal experts mean by electrons

Against Materialism 17

Moreover the existence of conscious intentional content is perfectly compatshyible with the existence of an externalist dimension of thought content-though not with the view that all content is external If as in Putnams famous example a person is thinking about earthly water at a time prior to the discovery of its chemical composition there is no reason to deny that they are in a sense thinking about H 20 But in such a situation the aspect of being about H 20 will obviously not be part of their conscious internal grasp of what they are thinking about in the way that the more superficial aspects of water will 16 And in a somewhat parallel way the person in Burges famous example who thinks he has arthritis in his thigh and to whom our standard belief-ascription practices ascribe a belief about arthritis (where this is among other things a disease that only occurs in joints) obviously does not consciously grasp the disease that he is thinking about in a way that involves this specific feature of itY But it is nonetheless impossible to describe either example in a convincing way without presupposing that the people in question do have something consciously in mind a substance having the superficial properties of water in Putnams example and a disease having certain fairly specific features in Burges Thus while it is possible to dispute the relative importance ofconscious internal thought content and external thought content of which the subject is not conscious examples of this sort provide no basis at all for denying that conscious internal content exists

The issue I want to raise here is whether a materialist view can account for the sort of conscious intentional content just characterized Can it account for conscious thoughts being about various things in a way that can be grasped or understood by the person in question In a way the answer has already been given Since materialist views really take no account at all of consciousness they obviously offer no account of this particular aspect of it But investigating this narrower aspect of the issue can still help to deepen the basic objection to

materialism Here it will be useful to bring the brilliant neurophysiologist Mary briefly

back onto the scene even though the black-and-white aspect of her situation is no longer relevant Suppose that Mary studies me as a subject and comes to

have a complete knowledge of my physical and neurophysiological makeup as I am thinking these various thoughts Can she determine on that basis what I am consciously thinking about at a particular moment

One thing that seems utterly clear is that she could not do this merely on the basis of knowing my internal physical characteristics-as it is sometimes put

I knowing everything physical that happens inside my skin There is no reason t f

is not the same thing as thinking about whatever is meant by electrons by whatever is meant by the societal experts

I 16 See Hilary Putnam (1975a) (Cambridge Cambridge Universiry Press) pp 215-71 17 See Tyler Burge (1979)

19 18 L Bonjour

at all to think that the internal structure of my physical and neurophysiological states could somehow by itself determine that I am thinking about weather tather than about the Middle East or the stock market

A functionalist would no doubt say that it is no surprise that Mary could not do this In order to know the complete causal or functional role of my internal states Mary also needs to know about their external causal relations to various things And it might be suggested if Mary knows all of the external causal relations in which my various states stand she will in fact be able ro figure out what I am consciously thinking about at any particular time No doubt the details that pick out any particular object of thought will be very complicated but there is it might be claimed no reason to doubt that in principle she could do this

Here we have a piece of materialist doctrine that again has a status very similar to that of a claim of theology It is obvious that no one ha~ even the beginnings of an idea of how to actually carry out an investigation that would yield a result of this kind-that the only rea~on for thinking that this could be done is the overriding assumption for which we have found no cogent basis that materialism must be true Among a multitude of other difficulties Mary would have to be able to figure out the content of thoughts that are confused or inaccurate or thoughts about imaginary or fictional entities or supernatural entities It is to say the least very hard to see how she could do this on the basis of a knowledge of causal relations to more ordinary sorts of things

But the problem for materialism is in fact even worse than that For as already emphasized it is an undeniable fact about conscious intentional content that I am able for the most part to consciously understand or be aware of what I am thinking about from the inside Clearly I do not in general do this on the basis of external causal knowledge I do not have such knowledge and would not know what to do about it if I did All that I normally have any sort of direct access to if materialism is true is my own internal physical and physiological states and thus my conscious understanding of what I am thinking about at a particular moment must be somehow a feature or result of those internal states alone Causal relations to external things may help to produce the relevant features of the internal states in question but there is no apparent way in which such external relations can somehow be partly constitutive of the fact that my conscious thoughts are about various things in a way of which I can be immediately aware But if these internal states are sufficient to fix the object of my thought in a way that is accessible to my understanding or awareness then knowing about those internal states should be sufficient for Mary a~ well without any knowledge of the external causal relations And yet as we have already seen it seems obvious that this is not the ca~e18

18 Ir is worth noting that the same thing is really true in the ase of qualia as well A persons awareness of one color rather than another when he or she looks at newly mown grass obviously

Against Materialism

Thus we have the basis for an argument that is parallel to Jacksons original argument about qualia Mary knows all the relevant physical facts she is not able on the basis of this knowledge to know what I am consciously thinking about at a particular moment but what I am thinking about at that moment is as surely a fact about the world as anything else therefore complete physical knowledge is not complete knowledge and so materialism is false

One way to further elaborate this point is to consider how it applies to

what is perhaps the most widely held materialist view of intentional content the view popularized by Jerry Fodor and many others that intentional mental states employ an internal language a language of thought19 Fodor calls this view the representational theory of the mind though it might better be called the symbolic theory of the mind For the crucial feature of the view is that the language of thought like any language is composed of symbols items that do not stand for anything by virtue of their intrinsic properties but whose representative character depends instead on the relations in which they stand to other things-for Fodor the sorts of causal relations that are captured in the idea of a causal or functional role1o Just as the word dog could in principle have stood for anything (or nothing at all) and in fact stands for a kind ofanimal rather than something else only because of causal relations that arise from the way it is used so also the symbols in the language of thought stand for whatever they stand for only by virrue of analogous sorts of relations and not in virtue of their intrinsic physical and neurophysiological properties Their intentional character is thus extrinsic not intrinsic

Proponents of the language ofthought rarely have much to say about conscious thoughts of the sort that we are focusing on here But it is clear that on their view what happens when I am consciously thinking about say the Middle East is that in some appropriate location in my overall cognitive operations there occurs a symbol (or set of symbols) that refers to the Middle East This symbol

f like the surrounding context in which it occurs is some neurophysiological state or some constellation of such states No one of course has at present any real

knowledge of the concrete nature of such symbols or their larger contexts but it f I will do no harm to follow Fodor in thinking of a mental blackboard on which

mental symbols are inscribed in appropriate ways Thus for me ro be consciously thinking about the Middle East is for me ro have the mental symbol that refers to

does not depend in a constitutive way on external causal rdations even though it may be causally produced by them Thus in that case too a knowledge of the persons internal physical and neurophysiological states alone should enable Mary to pick out One color rather than the other as the right one But it is even more obvious than in the original ase that this is not so

See eg Jerry Fodor Propositional Attitudes and Methodological Solipsism Considered as a Research Srrategy in Cognitive Science both reprinted in his 198 book Rtpwmll1rions (Cambridge MA MIT Press)

20 See for example Jerry Fodor chapter 4 Fodor has subsequently refined this view in various ways but none that affect issues being raised here

20 L Bonjour Against Materialism 21

the Middle East inscribed in the right way on this blackboard But the symbols reference to the Middle East to repeat depends not on its intrinsic physical or neurophysiological character alone but also on the relations in which it stands to other such symbols and directly or indirectly to the external world

Suppose now that Mary is studying my cognitive operations Suppose that she has somehow isolated what amounts to my mental blackboard and the various symbols written on it Obviously this will not in itself tell her what I am thinking about Even if she could somehow focus onthe specific symbol that refers to the Middle East and tell that it is functioning in a way that determines the object of my conscious thought (even though there is no reason to think that she could in fact do these things) she will not on this basis alone be able to tell what it is that this symbol in fact refers to Nor is there any plausibiliry to the idea that Mary could figure out the reference or meaning of the various mental symbols simply by examining their internal relations (0 each other21 Thus she will need once again to appeal to external causal relations ofvarious sorts

Bur how then am I able to be aware ofor understand from the inside what I am thinking about Once again I have no knowledge of those external relations (and would be very unlikely to be able to figure anything out from them even if I did) All that I plausibly have access to is the mental symbol or symbols and the surrounding system of states and this is apparently nor enough to determine the object of my thought

The only very obvious recourse here for the proponent of a language of thought is to construe my understanding or awareness of what I am thinking about disquotationally in relation to the language of thought When thinking about the Middle East I do so by using some mental symbol And when [ understand or am aware of what I am thinking about it might be suggested I in effect use that very same symbol what I am aware of is that I am thinking about the Middle East (whatever that is-that is whatever that symbol in fact refers (0) If the symbol in question did succeed in referring to the Middle East then this specification of what I am thinking about will refer to the Middle East as well and so will be correct But it is intuitively as obvious as anything could be that my awareness ofwhat I am thinking about normally involves more than this involves actually understanding (at some level of precision detail etc) what the Middle East is in a way that goes beyond merely repeating the same symbol Assuming for the moment that there really is a language of thought I understand my language of thought in a way parallel to the way in which I understand my own public language-and not in the merely disquotational way that could just as well be applied to a language of which I have no understanding at all

Here a proponent of the language of thought may want to reply that the difference in the public language case is merely that one language is a language

2 Such an idea has sometimes at least apparently been suggested For more discussion see 01)

19()8 book In Difmse ofPure Reason pp 174-80

that I successfully use-and that the same is true of my language of thought On this view the intuition that I understand what I am thinking about-or what I am talking about-in any stronger sense one that is not merely disquotational is merely an illusion But here again we have a view that it seems to me would appeal to no one who was not motivated by the conviction that materialism must be true

My conclusion is that the language of thought view has nothing useful to say about the most obvious sort of intentional content the intentional content that is involved in having something explicitly and consciously in mind Nor do I know of any other materialist account that does any better in this regard There is perhaps room for dispute about just how important conscious intentional content is in relation to the causation and explanation of behavior but no plausible way (0 deny that it genuinely exists Thus with respect to intentional content as with the case of qualitative content materialism seems to be utterly bankrupt as a general account of mental states and to be held merely as an article of faith

5 WHAT IS THE ALTERNATIVE

The last two sections serve merely to strengthen and deepen the fundamental objection to materialism already offered in section 2 consciousness genuinely exists materialism can oller no account that explains consciousness (or of the specific varieties thereof) or shows it [0 be merely material in character therefore (at least in the absence of any strong antecedent argument or presumption in favor of materialism) the indicated conclusion is that materialism is false There is more in heaven and earth than is dreamt of in materialist philosophy

But what do I mean by more Here as I see it there is very little that can be said in our current state of knowledge so thar the main result is that we have very little understanding ofconsciousness-or given the arguably central role of consciousness of mentaliry in general

In the first place there is no clear way in which the objections that I have raised against materialism support the classical substance dualist position Positing a separate mental substance that is characterized in almost entirely negative terms does nothing very obvious to explain consciousness in general or qualitative and intentional content in particular As far as I can see the main appeal of substance dualism is that the account of the supposed mental or spiritual substances is fur too vague and sketchy to provide the basis for any very clear argument that such substances could not be the locus of consciousness But this negative point hardly counts as an argument in favor of such a view

The obvious alternative is properry dualism the view that human persons and perhaps other kinds of animals have non-material or non-physical properties in addition to their physical ones with at least the main such properties being

16 L Bonjour

central things that having a dispositional belief that P disposes one to do A focus on belief in particular also has the unfortunate effect of making externalist accounts of intentional content seem more plausible than they possibly could if the emphasis were on conscious intentional states

For these reasons I will focus here explicitly on conscious thoughts As I sit writing this chapter a variety of conscious thoughts pass through my mind Many of these involve the assertion or endorsement of various propositions that materialism cannot account for consciousness that the trees outside my window are very bare that the weather looks cold and dank that the situation in the Middle East looks grim and so on Other thoughts are also propositional but in a way that does not involve assertion my conscious desire to get several pages written before lunch the hope that the stock market will continue to rise and so on It is doubtful that conscious thought must always be propositional in character but it will in any case simplify the issues to be discussed if we largely ignore the propositional aspect of these various thoughts and focus simply on their being conscious thoughts of or about various things or kinds of things materialism the trees the Middle East the stock market and so on

One crucial feature of such conscious thoughts is that when I have them I am in general consciously aware of or consciously understand or grasp what it is that I am thinking about (and also what I am thinking about it) When I think that the trees outside my window are bare I consciously understand that it is certain trees that I am thinking about (and along with this what sort of thing a tree is and which trees I have in mind) What exactly this conscious grasp of the object of thought involves varies from case to case and is sometimes not easy to precisely specify Moreover as will emerge it is something of which I think we presently have no real explanatory account of any substance But its existence is I submit completely undeniable Indeed being able in this way to consciously think about things to have them in mind is in many ways the most central and obvious feature of our mental lives

It is obvious that a persons conscious grasp of the object of their thought of what they are thinking about can vary on a number of dimensions it may be more or less precise more or less detailed more or less clear more or less complete But contrary to what is sometimes suggested it is rarely if ever merely disquotational in character Perhaps (though I doubt it) there are cases where a scientifically untutored person is thinking about eg electrons and where their sole grasp of what they are thinking is that it is what is referred to in their society or community by the word electron -so that what they are thinking about is in effect electrons (whatever they are) But this is surely not the ordinary situation when we think about various things 15

15 Notice that even a thought about what the relevant Ioci~tal ocp~rts mean by electrons would have to involve a non-disquotational element in the reference to those experts and also in the reference to the word to think about whatever it is that the societal experts mean by electrons

Against Materialism 17

Moreover the existence of conscious intentional content is perfectly compatshyible with the existence of an externalist dimension of thought content-though not with the view that all content is external If as in Putnams famous example a person is thinking about earthly water at a time prior to the discovery of its chemical composition there is no reason to deny that they are in a sense thinking about H 20 But in such a situation the aspect of being about H 20 will obviously not be part of their conscious internal grasp of what they are thinking about in the way that the more superficial aspects of water will 16 And in a somewhat parallel way the person in Burges famous example who thinks he has arthritis in his thigh and to whom our standard belief-ascription practices ascribe a belief about arthritis (where this is among other things a disease that only occurs in joints) obviously does not consciously grasp the disease that he is thinking about in a way that involves this specific feature of itY But it is nonetheless impossible to describe either example in a convincing way without presupposing that the people in question do have something consciously in mind a substance having the superficial properties of water in Putnams example and a disease having certain fairly specific features in Burges Thus while it is possible to dispute the relative importance ofconscious internal thought content and external thought content of which the subject is not conscious examples of this sort provide no basis at all for denying that conscious internal content exists

The issue I want to raise here is whether a materialist view can account for the sort of conscious intentional content just characterized Can it account for conscious thoughts being about various things in a way that can be grasped or understood by the person in question In a way the answer has already been given Since materialist views really take no account at all of consciousness they obviously offer no account of this particular aspect of it But investigating this narrower aspect of the issue can still help to deepen the basic objection to

materialism Here it will be useful to bring the brilliant neurophysiologist Mary briefly

back onto the scene even though the black-and-white aspect of her situation is no longer relevant Suppose that Mary studies me as a subject and comes to

have a complete knowledge of my physical and neurophysiological makeup as I am thinking these various thoughts Can she determine on that basis what I am consciously thinking about at a particular moment

One thing that seems utterly clear is that she could not do this merely on the basis of knowing my internal physical characteristics-as it is sometimes put

I knowing everything physical that happens inside my skin There is no reason t f

is not the same thing as thinking about whatever is meant by electrons by whatever is meant by the societal experts

I 16 See Hilary Putnam (1975a) (Cambridge Cambridge Universiry Press) pp 215-71 17 See Tyler Burge (1979)

19 18 L Bonjour

at all to think that the internal structure of my physical and neurophysiological states could somehow by itself determine that I am thinking about weather tather than about the Middle East or the stock market

A functionalist would no doubt say that it is no surprise that Mary could not do this In order to know the complete causal or functional role of my internal states Mary also needs to know about their external causal relations to various things And it might be suggested if Mary knows all of the external causal relations in which my various states stand she will in fact be able ro figure out what I am consciously thinking about at any particular time No doubt the details that pick out any particular object of thought will be very complicated but there is it might be claimed no reason to doubt that in principle she could do this

Here we have a piece of materialist doctrine that again has a status very similar to that of a claim of theology It is obvious that no one ha~ even the beginnings of an idea of how to actually carry out an investigation that would yield a result of this kind-that the only rea~on for thinking that this could be done is the overriding assumption for which we have found no cogent basis that materialism must be true Among a multitude of other difficulties Mary would have to be able to figure out the content of thoughts that are confused or inaccurate or thoughts about imaginary or fictional entities or supernatural entities It is to say the least very hard to see how she could do this on the basis of a knowledge of causal relations to more ordinary sorts of things

But the problem for materialism is in fact even worse than that For as already emphasized it is an undeniable fact about conscious intentional content that I am able for the most part to consciously understand or be aware of what I am thinking about from the inside Clearly I do not in general do this on the basis of external causal knowledge I do not have such knowledge and would not know what to do about it if I did All that I normally have any sort of direct access to if materialism is true is my own internal physical and physiological states and thus my conscious understanding of what I am thinking about at a particular moment must be somehow a feature or result of those internal states alone Causal relations to external things may help to produce the relevant features of the internal states in question but there is no apparent way in which such external relations can somehow be partly constitutive of the fact that my conscious thoughts are about various things in a way of which I can be immediately aware But if these internal states are sufficient to fix the object of my thought in a way that is accessible to my understanding or awareness then knowing about those internal states should be sufficient for Mary a~ well without any knowledge of the external causal relations And yet as we have already seen it seems obvious that this is not the ca~e18

18 Ir is worth noting that the same thing is really true in the ase of qualia as well A persons awareness of one color rather than another when he or she looks at newly mown grass obviously

Against Materialism

Thus we have the basis for an argument that is parallel to Jacksons original argument about qualia Mary knows all the relevant physical facts she is not able on the basis of this knowledge to know what I am consciously thinking about at a particular moment but what I am thinking about at that moment is as surely a fact about the world as anything else therefore complete physical knowledge is not complete knowledge and so materialism is false

One way to further elaborate this point is to consider how it applies to

what is perhaps the most widely held materialist view of intentional content the view popularized by Jerry Fodor and many others that intentional mental states employ an internal language a language of thought19 Fodor calls this view the representational theory of the mind though it might better be called the symbolic theory of the mind For the crucial feature of the view is that the language of thought like any language is composed of symbols items that do not stand for anything by virtue of their intrinsic properties but whose representative character depends instead on the relations in which they stand to other things-for Fodor the sorts of causal relations that are captured in the idea of a causal or functional role1o Just as the word dog could in principle have stood for anything (or nothing at all) and in fact stands for a kind ofanimal rather than something else only because of causal relations that arise from the way it is used so also the symbols in the language of thought stand for whatever they stand for only by virrue of analogous sorts of relations and not in virtue of their intrinsic physical and neurophysiological properties Their intentional character is thus extrinsic not intrinsic

Proponents of the language ofthought rarely have much to say about conscious thoughts of the sort that we are focusing on here But it is clear that on their view what happens when I am consciously thinking about say the Middle East is that in some appropriate location in my overall cognitive operations there occurs a symbol (or set of symbols) that refers to the Middle East This symbol

f like the surrounding context in which it occurs is some neurophysiological state or some constellation of such states No one of course has at present any real

knowledge of the concrete nature of such symbols or their larger contexts but it f I will do no harm to follow Fodor in thinking of a mental blackboard on which

mental symbols are inscribed in appropriate ways Thus for me ro be consciously thinking about the Middle East is for me ro have the mental symbol that refers to

does not depend in a constitutive way on external causal rdations even though it may be causally produced by them Thus in that case too a knowledge of the persons internal physical and neurophysiological states alone should enable Mary to pick out One color rather than the other as the right one But it is even more obvious than in the original ase that this is not so

See eg Jerry Fodor Propositional Attitudes and Methodological Solipsism Considered as a Research Srrategy in Cognitive Science both reprinted in his 198 book Rtpwmll1rions (Cambridge MA MIT Press)

20 See for example Jerry Fodor chapter 4 Fodor has subsequently refined this view in various ways but none that affect issues being raised here

20 L Bonjour Against Materialism 21

the Middle East inscribed in the right way on this blackboard But the symbols reference to the Middle East to repeat depends not on its intrinsic physical or neurophysiological character alone but also on the relations in which it stands to other such symbols and directly or indirectly to the external world

Suppose now that Mary is studying my cognitive operations Suppose that she has somehow isolated what amounts to my mental blackboard and the various symbols written on it Obviously this will not in itself tell her what I am thinking about Even if she could somehow focus onthe specific symbol that refers to the Middle East and tell that it is functioning in a way that determines the object of my conscious thought (even though there is no reason to think that she could in fact do these things) she will not on this basis alone be able to tell what it is that this symbol in fact refers to Nor is there any plausibiliry to the idea that Mary could figure out the reference or meaning of the various mental symbols simply by examining their internal relations (0 each other21 Thus she will need once again to appeal to external causal relations ofvarious sorts

Bur how then am I able to be aware ofor understand from the inside what I am thinking about Once again I have no knowledge of those external relations (and would be very unlikely to be able to figure anything out from them even if I did) All that I plausibly have access to is the mental symbol or symbols and the surrounding system of states and this is apparently nor enough to determine the object of my thought

The only very obvious recourse here for the proponent of a language of thought is to construe my understanding or awareness of what I am thinking about disquotationally in relation to the language of thought When thinking about the Middle East I do so by using some mental symbol And when [ understand or am aware of what I am thinking about it might be suggested I in effect use that very same symbol what I am aware of is that I am thinking about the Middle East (whatever that is-that is whatever that symbol in fact refers (0) If the symbol in question did succeed in referring to the Middle East then this specification of what I am thinking about will refer to the Middle East as well and so will be correct But it is intuitively as obvious as anything could be that my awareness ofwhat I am thinking about normally involves more than this involves actually understanding (at some level of precision detail etc) what the Middle East is in a way that goes beyond merely repeating the same symbol Assuming for the moment that there really is a language of thought I understand my language of thought in a way parallel to the way in which I understand my own public language-and not in the merely disquotational way that could just as well be applied to a language of which I have no understanding at all

Here a proponent of the language of thought may want to reply that the difference in the public language case is merely that one language is a language

2 Such an idea has sometimes at least apparently been suggested For more discussion see 01)

19()8 book In Difmse ofPure Reason pp 174-80

that I successfully use-and that the same is true of my language of thought On this view the intuition that I understand what I am thinking about-or what I am talking about-in any stronger sense one that is not merely disquotational is merely an illusion But here again we have a view that it seems to me would appeal to no one who was not motivated by the conviction that materialism must be true

My conclusion is that the language of thought view has nothing useful to say about the most obvious sort of intentional content the intentional content that is involved in having something explicitly and consciously in mind Nor do I know of any other materialist account that does any better in this regard There is perhaps room for dispute about just how important conscious intentional content is in relation to the causation and explanation of behavior but no plausible way (0 deny that it genuinely exists Thus with respect to intentional content as with the case of qualitative content materialism seems to be utterly bankrupt as a general account of mental states and to be held merely as an article of faith

5 WHAT IS THE ALTERNATIVE

The last two sections serve merely to strengthen and deepen the fundamental objection to materialism already offered in section 2 consciousness genuinely exists materialism can oller no account that explains consciousness (or of the specific varieties thereof) or shows it [0 be merely material in character therefore (at least in the absence of any strong antecedent argument or presumption in favor of materialism) the indicated conclusion is that materialism is false There is more in heaven and earth than is dreamt of in materialist philosophy

But what do I mean by more Here as I see it there is very little that can be said in our current state of knowledge so thar the main result is that we have very little understanding ofconsciousness-or given the arguably central role of consciousness of mentaliry in general

In the first place there is no clear way in which the objections that I have raised against materialism support the classical substance dualist position Positing a separate mental substance that is characterized in almost entirely negative terms does nothing very obvious to explain consciousness in general or qualitative and intentional content in particular As far as I can see the main appeal of substance dualism is that the account of the supposed mental or spiritual substances is fur too vague and sketchy to provide the basis for any very clear argument that such substances could not be the locus of consciousness But this negative point hardly counts as an argument in favor of such a view

The obvious alternative is properry dualism the view that human persons and perhaps other kinds of animals have non-material or non-physical properties in addition to their physical ones with at least the main such properties being

19 18 L Bonjour

at all to think that the internal structure of my physical and neurophysiological states could somehow by itself determine that I am thinking about weather tather than about the Middle East or the stock market

A functionalist would no doubt say that it is no surprise that Mary could not do this In order to know the complete causal or functional role of my internal states Mary also needs to know about their external causal relations to various things And it might be suggested if Mary knows all of the external causal relations in which my various states stand she will in fact be able ro figure out what I am consciously thinking about at any particular time No doubt the details that pick out any particular object of thought will be very complicated but there is it might be claimed no reason to doubt that in principle she could do this

Here we have a piece of materialist doctrine that again has a status very similar to that of a claim of theology It is obvious that no one ha~ even the beginnings of an idea of how to actually carry out an investigation that would yield a result of this kind-that the only rea~on for thinking that this could be done is the overriding assumption for which we have found no cogent basis that materialism must be true Among a multitude of other difficulties Mary would have to be able to figure out the content of thoughts that are confused or inaccurate or thoughts about imaginary or fictional entities or supernatural entities It is to say the least very hard to see how she could do this on the basis of a knowledge of causal relations to more ordinary sorts of things

But the problem for materialism is in fact even worse than that For as already emphasized it is an undeniable fact about conscious intentional content that I am able for the most part to consciously understand or be aware of what I am thinking about from the inside Clearly I do not in general do this on the basis of external causal knowledge I do not have such knowledge and would not know what to do about it if I did All that I normally have any sort of direct access to if materialism is true is my own internal physical and physiological states and thus my conscious understanding of what I am thinking about at a particular moment must be somehow a feature or result of those internal states alone Causal relations to external things may help to produce the relevant features of the internal states in question but there is no apparent way in which such external relations can somehow be partly constitutive of the fact that my conscious thoughts are about various things in a way of which I can be immediately aware But if these internal states are sufficient to fix the object of my thought in a way that is accessible to my understanding or awareness then knowing about those internal states should be sufficient for Mary a~ well without any knowledge of the external causal relations And yet as we have already seen it seems obvious that this is not the ca~e18

18 Ir is worth noting that the same thing is really true in the ase of qualia as well A persons awareness of one color rather than another when he or she looks at newly mown grass obviously

Against Materialism

Thus we have the basis for an argument that is parallel to Jacksons original argument about qualia Mary knows all the relevant physical facts she is not able on the basis of this knowledge to know what I am consciously thinking about at a particular moment but what I am thinking about at that moment is as surely a fact about the world as anything else therefore complete physical knowledge is not complete knowledge and so materialism is false

One way to further elaborate this point is to consider how it applies to

what is perhaps the most widely held materialist view of intentional content the view popularized by Jerry Fodor and many others that intentional mental states employ an internal language a language of thought19 Fodor calls this view the representational theory of the mind though it might better be called the symbolic theory of the mind For the crucial feature of the view is that the language of thought like any language is composed of symbols items that do not stand for anything by virtue of their intrinsic properties but whose representative character depends instead on the relations in which they stand to other things-for Fodor the sorts of causal relations that are captured in the idea of a causal or functional role1o Just as the word dog could in principle have stood for anything (or nothing at all) and in fact stands for a kind ofanimal rather than something else only because of causal relations that arise from the way it is used so also the symbols in the language of thought stand for whatever they stand for only by virrue of analogous sorts of relations and not in virtue of their intrinsic physical and neurophysiological properties Their intentional character is thus extrinsic not intrinsic

Proponents of the language ofthought rarely have much to say about conscious thoughts of the sort that we are focusing on here But it is clear that on their view what happens when I am consciously thinking about say the Middle East is that in some appropriate location in my overall cognitive operations there occurs a symbol (or set of symbols) that refers to the Middle East This symbol

f like the surrounding context in which it occurs is some neurophysiological state or some constellation of such states No one of course has at present any real

knowledge of the concrete nature of such symbols or their larger contexts but it f I will do no harm to follow Fodor in thinking of a mental blackboard on which

mental symbols are inscribed in appropriate ways Thus for me ro be consciously thinking about the Middle East is for me ro have the mental symbol that refers to

does not depend in a constitutive way on external causal rdations even though it may be causally produced by them Thus in that case too a knowledge of the persons internal physical and neurophysiological states alone should enable Mary to pick out One color rather than the other as the right one But it is even more obvious than in the original ase that this is not so

See eg Jerry Fodor Propositional Attitudes and Methodological Solipsism Considered as a Research Srrategy in Cognitive Science both reprinted in his 198 book Rtpwmll1rions (Cambridge MA MIT Press)

20 See for example Jerry Fodor chapter 4 Fodor has subsequently refined this view in various ways but none that affect issues being raised here

20 L Bonjour Against Materialism 21

the Middle East inscribed in the right way on this blackboard But the symbols reference to the Middle East to repeat depends not on its intrinsic physical or neurophysiological character alone but also on the relations in which it stands to other such symbols and directly or indirectly to the external world

Suppose now that Mary is studying my cognitive operations Suppose that she has somehow isolated what amounts to my mental blackboard and the various symbols written on it Obviously this will not in itself tell her what I am thinking about Even if she could somehow focus onthe specific symbol that refers to the Middle East and tell that it is functioning in a way that determines the object of my conscious thought (even though there is no reason to think that she could in fact do these things) she will not on this basis alone be able to tell what it is that this symbol in fact refers to Nor is there any plausibiliry to the idea that Mary could figure out the reference or meaning of the various mental symbols simply by examining their internal relations (0 each other21 Thus she will need once again to appeal to external causal relations ofvarious sorts

Bur how then am I able to be aware ofor understand from the inside what I am thinking about Once again I have no knowledge of those external relations (and would be very unlikely to be able to figure anything out from them even if I did) All that I plausibly have access to is the mental symbol or symbols and the surrounding system of states and this is apparently nor enough to determine the object of my thought

The only very obvious recourse here for the proponent of a language of thought is to construe my understanding or awareness of what I am thinking about disquotationally in relation to the language of thought When thinking about the Middle East I do so by using some mental symbol And when [ understand or am aware of what I am thinking about it might be suggested I in effect use that very same symbol what I am aware of is that I am thinking about the Middle East (whatever that is-that is whatever that symbol in fact refers (0) If the symbol in question did succeed in referring to the Middle East then this specification of what I am thinking about will refer to the Middle East as well and so will be correct But it is intuitively as obvious as anything could be that my awareness ofwhat I am thinking about normally involves more than this involves actually understanding (at some level of precision detail etc) what the Middle East is in a way that goes beyond merely repeating the same symbol Assuming for the moment that there really is a language of thought I understand my language of thought in a way parallel to the way in which I understand my own public language-and not in the merely disquotational way that could just as well be applied to a language of which I have no understanding at all

Here a proponent of the language of thought may want to reply that the difference in the public language case is merely that one language is a language

2 Such an idea has sometimes at least apparently been suggested For more discussion see 01)

19()8 book In Difmse ofPure Reason pp 174-80

that I successfully use-and that the same is true of my language of thought On this view the intuition that I understand what I am thinking about-or what I am talking about-in any stronger sense one that is not merely disquotational is merely an illusion But here again we have a view that it seems to me would appeal to no one who was not motivated by the conviction that materialism must be true

My conclusion is that the language of thought view has nothing useful to say about the most obvious sort of intentional content the intentional content that is involved in having something explicitly and consciously in mind Nor do I know of any other materialist account that does any better in this regard There is perhaps room for dispute about just how important conscious intentional content is in relation to the causation and explanation of behavior but no plausible way (0 deny that it genuinely exists Thus with respect to intentional content as with the case of qualitative content materialism seems to be utterly bankrupt as a general account of mental states and to be held merely as an article of faith

5 WHAT IS THE ALTERNATIVE

The last two sections serve merely to strengthen and deepen the fundamental objection to materialism already offered in section 2 consciousness genuinely exists materialism can oller no account that explains consciousness (or of the specific varieties thereof) or shows it [0 be merely material in character therefore (at least in the absence of any strong antecedent argument or presumption in favor of materialism) the indicated conclusion is that materialism is false There is more in heaven and earth than is dreamt of in materialist philosophy

But what do I mean by more Here as I see it there is very little that can be said in our current state of knowledge so thar the main result is that we have very little understanding ofconsciousness-or given the arguably central role of consciousness of mentaliry in general

In the first place there is no clear way in which the objections that I have raised against materialism support the classical substance dualist position Positing a separate mental substance that is characterized in almost entirely negative terms does nothing very obvious to explain consciousness in general or qualitative and intentional content in particular As far as I can see the main appeal of substance dualism is that the account of the supposed mental or spiritual substances is fur too vague and sketchy to provide the basis for any very clear argument that such substances could not be the locus of consciousness But this negative point hardly counts as an argument in favor of such a view

The obvious alternative is properry dualism the view that human persons and perhaps other kinds of animals have non-material or non-physical properties in addition to their physical ones with at least the main such properties being

20 L Bonjour Against Materialism 21

the Middle East inscribed in the right way on this blackboard But the symbols reference to the Middle East to repeat depends not on its intrinsic physical or neurophysiological character alone but also on the relations in which it stands to other such symbols and directly or indirectly to the external world

Suppose now that Mary is studying my cognitive operations Suppose that she has somehow isolated what amounts to my mental blackboard and the various symbols written on it Obviously this will not in itself tell her what I am thinking about Even if she could somehow focus onthe specific symbol that refers to the Middle East and tell that it is functioning in a way that determines the object of my conscious thought (even though there is no reason to think that she could in fact do these things) she will not on this basis alone be able to tell what it is that this symbol in fact refers to Nor is there any plausibiliry to the idea that Mary could figure out the reference or meaning of the various mental symbols simply by examining their internal relations (0 each other21 Thus she will need once again to appeal to external causal relations ofvarious sorts

Bur how then am I able to be aware ofor understand from the inside what I am thinking about Once again I have no knowledge of those external relations (and would be very unlikely to be able to figure anything out from them even if I did) All that I plausibly have access to is the mental symbol or symbols and the surrounding system of states and this is apparently nor enough to determine the object of my thought

The only very obvious recourse here for the proponent of a language of thought is to construe my understanding or awareness of what I am thinking about disquotationally in relation to the language of thought When thinking about the Middle East I do so by using some mental symbol And when [ understand or am aware of what I am thinking about it might be suggested I in effect use that very same symbol what I am aware of is that I am thinking about the Middle East (whatever that is-that is whatever that symbol in fact refers (0) If the symbol in question did succeed in referring to the Middle East then this specification of what I am thinking about will refer to the Middle East as well and so will be correct But it is intuitively as obvious as anything could be that my awareness ofwhat I am thinking about normally involves more than this involves actually understanding (at some level of precision detail etc) what the Middle East is in a way that goes beyond merely repeating the same symbol Assuming for the moment that there really is a language of thought I understand my language of thought in a way parallel to the way in which I understand my own public language-and not in the merely disquotational way that could just as well be applied to a language of which I have no understanding at all

Here a proponent of the language of thought may want to reply that the difference in the public language case is merely that one language is a language

2 Such an idea has sometimes at least apparently been suggested For more discussion see 01)

19()8 book In Difmse ofPure Reason pp 174-80

that I successfully use-and that the same is true of my language of thought On this view the intuition that I understand what I am thinking about-or what I am talking about-in any stronger sense one that is not merely disquotational is merely an illusion But here again we have a view that it seems to me would appeal to no one who was not motivated by the conviction that materialism must be true

My conclusion is that the language of thought view has nothing useful to say about the most obvious sort of intentional content the intentional content that is involved in having something explicitly and consciously in mind Nor do I know of any other materialist account that does any better in this regard There is perhaps room for dispute about just how important conscious intentional content is in relation to the causation and explanation of behavior but no plausible way (0 deny that it genuinely exists Thus with respect to intentional content as with the case of qualitative content materialism seems to be utterly bankrupt as a general account of mental states and to be held merely as an article of faith

5 WHAT IS THE ALTERNATIVE

The last two sections serve merely to strengthen and deepen the fundamental objection to materialism already offered in section 2 consciousness genuinely exists materialism can oller no account that explains consciousness (or of the specific varieties thereof) or shows it [0 be merely material in character therefore (at least in the absence of any strong antecedent argument or presumption in favor of materialism) the indicated conclusion is that materialism is false There is more in heaven and earth than is dreamt of in materialist philosophy

But what do I mean by more Here as I see it there is very little that can be said in our current state of knowledge so thar the main result is that we have very little understanding ofconsciousness-or given the arguably central role of consciousness of mentaliry in general

In the first place there is no clear way in which the objections that I have raised against materialism support the classical substance dualist position Positing a separate mental substance that is characterized in almost entirely negative terms does nothing very obvious to explain consciousness in general or qualitative and intentional content in particular As far as I can see the main appeal of substance dualism is that the account of the supposed mental or spiritual substances is fur too vague and sketchy to provide the basis for any very clear argument that such substances could not be the locus of consciousness But this negative point hardly counts as an argument in favor of such a view

The obvious alternative is properry dualism the view that human persons and perhaps other kinds of animals have non-material or non-physical properties in addition to their physical ones with at least the main such properties being

22 23 L Bonjour

the various kinds of consciousness including the central ones that have been discussed here In away this view seems obviously correct The properties in question genuinely exist and seem on the basis of the filliure of materialism to

explain or account for them to be clearly non-material in character Bur without some further explanation of what such properties amount to or of how they could be properties of a mostly material organism-or for that matter of an immaterial subsrance-the property dualist view yields little in the way of real understanding and hardly counts as a serious account of the nature of mental states

One somewhat more definite result can I think be derived from the discussion of conscious intentional thought If when I think consciously about things I am able to know what it is that I am thinking about without knowing anything further about external relations then what the states in question are about must apparently be an intrinsic feature of them they must have intrinsic intentionality as opposed to an intentionality like that of language (including a language of thought) that is derived from external relations When I am consciously thinking about say trees there must be something about the intrinsic character of my state of mind that makes it about trees (and in a way that is immediately apparent to me) Here we have a conclusion that vety few would accept and that many would regard as virtually absurd All I can say is that it seems to me clearly required by the facts of the situation

But how could the intrinsic character of a state definitively pick our something external to it in this way I do not claim to have anything like a clear answer to

this question but I will indulge in a bit of what seems to me initially plausible speculation First I offer the surmise that what is needed to account for intrinsic intentionality in general is an account of two sorts of intrinsically intentional elements first intrinsic reference to properties of various kinds and second intrinsically indexical content

About the latter of these it is reasonably plausible to suppose that indexical content of all kinds can be reduced to an indexical reference to the self with other things including other places and times being indexically specified by appeal to their relations to the self Such a view has sometimes been suggested by others as well22 but I have no space to develop it further here

Intrinsic reference to properties seems more difficult Including anything in a state that merely in some way stands for or represents a property does not seem to yield intrinsic intentionality since the reference to the target property will also depend on the external relation between this representing element and that property itself Having a symbolic element that stands for the target property in question obviously will not work for reasons that we have already seen in the earlier discussion But having a representing element that resembles the target property also seems inadequate If the representing element resembles the target

22 See for example David Lewis (1979)

Against Materiaism

property by having some other distinct property then the connection to the target property seems to depend on the relation of resemblance in a way that makes the reference to the target property no longer intrinsic It is also hard to see how someone who has direct access only to the resembling property would be able to be aware that they were thinking of the target ptoperty (or of something else that wa~ picked out by appeal to the target property) Such a person would seemingly have only the resembling property and not the target property explicitly in mind

I

Thus what seems to be required is that the intrinsically intentional state actually involve in some way the target property itself Nothing else seems adequate to make the reference to that property both intrinsic and in principle accessible to the person having the thought Obviously though this cannot in general involve the intrinsically intentional state or some component of it literally instantiating the target property for obviously we can think about lots of properties that are not literally instantiated in our intentional states Elsewhere I have speculated that what might be involved is the state or some component instantiating a complex universal that has the target property as an ingredient in some appropriate way23 But while this proposal seems to have in a way the right

I sort of structure I do not really claim to have even an initial understanding of what it would involve or how it would work

My conclusion remains almost entirely negative We can see that consciousness exists and we can see what this specific sort of consciousness in particular would

i have to involve-namely intrinsic intentionality And seeing what intrinsic intentionality in turn would require makes it if anything even clearer that there is no reason at all to think that a merely material state could have this characteristic But how consciousness in general or intrinsic intentionality inI particular can be explained and accounted for is something about which if 1 am right we know almost nothing

1 See BonJour (1998 IBO-61